"""演示种子数据 — 按 5 个业务场景生成完整示例数据。 覆盖模型：Tenant / User / Company / MonthlyReport / HealthScore / RiskEvent / WeakSignal / Task / OKR / MilestoneTree / SynergyOpportunity / DecisionSentinel / NudgeRecord / MajorEvent / InvestmentAgreement / BoardMeeting / FinancialData / Hypothesis / AARRecord / TeamMember / TalentProfile / AuditLog。 运行方式： cd backend && python -m scripts.seed_demo_data """ import asyncio import json import random from datetime import datetime, timedelta, timezone from sqlalchemy import text from sqlalchemy.ext.asyncio import AsyncSession from app.core.database import async_session_factory, engine, Base from app.core.security import hash_password from app.models import ( AARRecord, AgentExecution, AuditLog, BoardMeeting, Company, CompanyFundLink, CustomerAcquisitionPlan, DataSource, DecisionSentinel, DigitalTwinModel, EvaluationTemplate, ExitPrediction, FinancialData, FirmProfile, Fund, FundProfile, HealthScore, Hypothesis, InquiryList, InterventionEvent, InterventionResult, InvestmentAgreement, KnowledgeChunk, KnowledgeNode, MajorEvent, ManagerProfile, MilestoneTree, MonteCarloSimulation, MonthlyReport, NudgeRecord, OKR, PeerLearningCircle, PortfolioRebalancing, PreMortemRecord, ProductDiagnostic, RedTeamRecord, RiskEvent, SynergyOpportunity, Task, TalentProfile, TeamMember, Tenant, User, WeakSignal, ) UTC = timezone.utc # ─── 固定 ID（方便跨表引用） ────────────────────────────────────────── TENANT_ID = "00000000-0000-0000-0000-000000000001" ADMIN_ID = "00000000-0000-0000-0000-000000000002" GP_ID = "00000000-0000-0000-0000-000000000003" LEAD_ID = "00000000-0000-0000-0000-000000000004" MGR_ID = "00000000-0000-0000-0000-000000000005" FOUNDER_A = "00000000-0000-0000-0000-000000000006" FOUNDER_B = "00000000-0000-0000-0000-000000000007" FOUNDER_C = "00000000-0000-0000-0000-000000000008" COMP_A = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaa1" COMP_B = "bbbbbbbb-bbbb-bbbb-bbbb-bbbbbbbbbbb2" COMP_C = "cccccccc-cccc-cccc-cccc-ccccccccccc3" COMP_D = "dddddddd-dddd-dddd-dddd-ddddddddddd4" COMP_E = "eeeeeeee-eeee-eeee-eeee-eeeeeeeeeee5" NOW = datetime.now(UTC) D = lambda days: NOW - timedelta(days=days) # noqa: E731 async def seed() -> None: """生成全部演示数据。""" async with engine.begin() as conn: await conn.run_sync(Base.metadata.drop_all) await conn.run_sync(Base.metadata.create_all) async with async_session_factory() as db: await _seed_tenant_users(db) await _seed_companies(db) await _seed_funds(db) await _seed_profiles(db) await _seed_evaluation_templates(db) await _seed_financial(db) await _seed_reports(db) await _seed_inquiries(db) await _seed_health(db) await _seed_health_history(db) await _seed_risks(db) await _seed_weak_signals(db) await _seed_tasks(db) await _seed_okrs(db) await _seed_milestones(db) await _seed_synergies(db) await _seed_sentinels(db) await _seed_nudges(db) await _seed_events(db) await _seed_agreements(db) await _seed_boards(db) await _seed_hypotheses(db) await _seed_aars(db) await _seed_team_members(db) await _seed_talents(db) await _seed_agent_executions(db) await _seed_customer_plans(db) await _seed_exit_predictions(db) await _seed_pre_mortems(db) await _seed_product_diagnostics(db) await _seed_peer_circles(db) await _seed_digital_twins(db) await _seed_interventions(db) await _seed_portfolio_simulations(db) await _seed_knowledge_graph(db) await _seed_data_sources(db) await _seed_audit_logs(db) await db.commit() print("✅ 演示数据生成完成") print(f" 租户: {TENANT_ID}") print(f" 用户: admin/gp/lead/mgr/founder_a/founder_b/founder_c") print(f" 企业: 智链科技/云栈数据/深瞳智能/量子芯微/光合生物") print(" 登录密码均为: demo123456") # ─── 租户 + 用户 ─────────────────────────────────────────────────────── async def _seed_tenant_users(db: AsyncSession) -> None: tenant = Tenant( id=TENANT_ID, name="远见资本", type="vc", config_json={"fund_name": "远见三期", "fund_size": "5亿人民币"}, ) db.add(tenant) users = [ User(id=ADMIN_ID, tenant_id=TENANT_ID, email="admin@demo.com", password_hash=hash_password("demo123456"), name="系统管理员", role="admin", is_active=True), User(id=GP_ID, tenant_id=TENANT_ID, email="gp@demo.com", password_hash=hash_password("demo123456"), name="张远见", role="gp", is_active=True), User(id=LEAD_ID, tenant_id=TENANT_ID, email="lead@demo.com", password_hash=hash_password("demo123456"), name="李投后", role="post_invest_lead", is_active=True), User(id=MGR_ID, tenant_id=TENANT_ID, email="mgr@demo.com", password_hash=hash_password("demo123456"), name="王经理", role="investor", is_active=True), User(id=FOUNDER_A, tenant_id=TENANT_ID, email="founder_a@demo.com", password_hash=hash_password("demo123456"), name="陈智链", role="founder", is_active=True), User(id=FOUNDER_B, tenant_id=TENANT_ID, email="founder_b@demo.com", password_hash=hash_password("demo123456"), name="刘云栈", role="founder", is_active=True), User(id=FOUNDER_C, tenant_id=TENANT_ID, email="founder_c@demo.com", password_hash=hash_password("demo123456"), name="赵深瞳", role="founder", is_active=True), ] for u in users: db.add(u) await db.flush() print(" ✓ 租户 + 7 用户") # ─── 企业 ────────────────────────────────────────────────────────────── async def _seed_companies(db: AsyncSession) -> None: companies = [ Company(id=COMP_A, tenant_id=TENANT_ID, name="智链科技有限公司", industry="AI/供应链", stage="b", description="利用大模型优化供应链决策，服务制造业客户。", founded_at=datetime(2021, 3, 1, tzinfo=UTC), total_funding="1.2亿人民币", website="https://zhilian.example.com"), Company(id=COMP_B, tenant_id=TENANT_ID, name="云栈数据科技", industry="AI/数据平台", stage="a", description="企业级数据治理与 AI 训练数据管理平台。", founded_at=datetime(2022, 1, 1, tzinfo=UTC), total_funding="5000万人民币", website="https://yunzhan.example.com"), Company(id=COMP_C, tenant_id=TENANT_ID, name="深瞳智能", industry="AI/计算机视觉", stage="b", description="工业质检视觉 AI，服务半导体和新能源行业。", founded_at=datetime(2020, 6, 1, tzinfo=UTC), total_funding="2亿人民币", website="https://shentong.example.com"), Company(id=COMP_D, tenant_id=TENANT_ID, name="量子芯微电子", industry="芯片/AI算力", stage="seed", description="AI 推理芯片设计，边缘端低功耗 NPU。", founded_at=datetime(2023, 9, 1, tzinfo=UTC), total_funding="1500万人民币", website="https://quantumchip.example.com"), Company(id=COMP_E, tenant_id=TENANT_ID, name="光合生物医疗", industry="AI/医疗", stage="a", description="AI 辅助药物发现，聚焦小分子靶点筛选。", founded_at=datetime(2022, 9, 1, tzinfo=UTC), total_funding="8000万人民币", website="https://guanghe.example.com"), ] for c in companies: db.add(c) await db.flush() print(" ✓ 5 企业") # ─── 财务数据 ────────────────────────────────────────────────────────── async def _seed_financial(db: AsyncSession) -> None: """每家企业生成近 6 个月利润表 + 现金流。""" for comp_id, revenue, burn, cash in [ (COMP_A, 800, 350, 4200), # 月营收800万, 消耗350万, 现金4200万 (COMP_B, 150, 120, 1800), (COMP_C, 1200, 480, 6500), (COMP_D, 0, 80, 600), (COMP_E, 80, 200, 2400), ]: for i in range(6): y, m = (NOW.year, NOW.month - i) if NOW.month - i > 0 else (NOW.year - 1, 12 + NOW.month - i) # 营收逐月递减（历史数据），展示增长趋势 hist_revenue = int(revenue * (1 - 0.03 * i)) if revenue > 0 else 0 db.add(FinancialData( company_id=comp_id, period_year=y, period_month=m, statement_type="income", data_json={"revenue": hist_revenue, "cogs": int(hist_revenue * 0.4), "gross_profit": int(hist_revenue * 0.6), "opex": burn, "net_income": hist_revenue - burn - int(hist_revenue * 0.4)}, source="monthly_report", credibility_score=85.0 - i * 5, )) db.add(FinancialData( company_id=comp_id, period_year=y, period_month=m, statement_type="cash_flow", data_json={"operating_cf": hist_revenue - burn, "investing_cf": -50, "financing_cf": 0, "net_cf": hist_revenue - burn - 50, "cash_balance": cash - i * burn}, source="monthly_report", credibility_score=85.0 - i * 5, )) await db.flush() print(" ✓ 财务数据 (5 企业 x 6 月)") # ─── 月报 ────────────────────────────────────────────────────────────── async def _seed_reports(db: AsyncSession) -> None: reports = [ # 智链科技 — 已提交 + AI 解析 MonthlyReport(company_id=COMP_A, period_year=NOW.year, period_month=NOW.month - 1 if NOW.month > 1 else 12, status="ai_parsed", submitted_by=FOUNDER_A, submitted_at=D(10), raw_content="本月营收800万，环比增长15%。新签客户3家，流失1家。团队新增5人。AI模型推理成本降低12%。", structured_data={"revenue": 800, "mom_growth": 0.15, "new_customers": 3, "churned": 1, "headcount": 45, "ai_cost_reduction": 0.12}, ai_summary="智链科技本月表现良好，营收增长15%，AI推理成本优化显著。客户净增2家，团队扩张至45人。", ai_concerns=[{"level": "medium", "item": "客户流失率环比上升", "detail": "本月流失1家年合同客户"}, {"level": "low", "item": "现金Runway 12个月", "detail": "按当前消耗速度可维持12个月"}]), # 云栈数据 — 已提交 MonthlyReport(company_id=COMP_B, period_year=NOW.year, period_month=NOW.month - 1 if NOW.month > 1 else 12, status="ai_parsed", submitted_by=FOUNDER_B, submitted_at=D(8), raw_content="本月营收150万，环比持平。完成数据治理平台v2.0发布。新签1家金融客户。", structured_data={"revenue": 150, "mom_growth": 0.0, "new_customers": 1, "churned": 0, "headcount": 22}, ai_summary="云栈数据营收持平，产品迭代正常。金融客户拓展是亮点。", ai_concerns=[{"level": "high", "item": "营收增长停滞", "detail": "连续2个月环比0%增长"}, {"level": "medium", "item": "现金Runway 15个月", "detail": "需关注融资节奏"}]), # 深瞳智能 — 已提交 MonthlyReport(company_id=COMP_C, period_year=NOW.year, period_month=NOW.month - 1 if NOW.month > 1 else 12, status="reviewed", submitted_by=FOUNDER_C, submitted_at=D(12), reviewed_by=MGR_ID, reviewed_at=D(5), raw_content="本月营收1200万，环比增长22%。半导体客户占比提升至60%。新增专利2项。", structured_data={"revenue": 1200, "mom_growth": 0.22, "new_customers": 2, "churned": 0, "headcount": 68, "patents": 2}, ai_summary="深瞳智能强劲增长，半导体行业渗透加深。专利积累构建技术壁垒。", ai_concerns=[{"level": "low", "item": "客户集中度风险", "detail": "半导体行业占比60%"}]), # 量子芯微 — 草稿 MonthlyReport(company_id=COMP_D, period_year=NOW.year, period_month=NOW.month - 1 if NOW.month > 1 else 12, status="draft", submitted_by=None, raw_content="", structured_data=None, ai_summary=None, ai_concerns=None), # 光合生物 — 延迟提交 MonthlyReport(company_id=COMP_E, period_year=NOW.year, period_month=NOW.month - 2 if NOW.month > 2 else 11, status="ai_parsed", submitted_by=FOUNDER_A, submitted_at=D(40), raw_content="本月营收80万，来自2个合作药企里程碑付款。研发进展：3个靶点进入先导化合物阶段。", structured_data={"revenue": 80, "milestone_payment": 80, "targets": 3, "headcount": 18}, ai_summary="光合生物收入来自里程碑付款，研发管线推进正常。", ai_concerns=[{"level": "high", "item": "月报延迟提交", "detail": "延迟20天"}, {"level": "medium", "item": "现金消耗快", "detail": "月消耗200万，Runway 12月"}]), ] for r in reports: db.add(r) await db.flush() # 历史月报 — 每家企业再补 5 个月，展示趋势 historical_data = [ # (company_id, founder_id, months_ago, revenue, growth, new_cust, churned, headcount, summary, concerns) (COMP_A, FOUNDER_A, 2, 696, 0.08, 2, 0, 40, "营收稳步增长，新签2家客户。", []), (COMP_A, FOUNDER_A, 3, 644, 0.05, 1, 1, 38, "营收增长5%，客户流失1家。", [{"level": "low", "item": "客户流失", "detail": "流失1家小客户"}]), (COMP_A, FOUNDER_A, 4, 613, 0.03, 1, 0, 37, "营收小幅增长，团队稳定。", []), (COMP_A, FOUNDER_A, 5, 595, 0.02, 0, 0, 35, "营收微增，无新签客户。", [{"level": "medium", "item": "增长放缓", "detail": "环比仅2%"}]), (COMP_A, FOUNDER_A, 6, 583, 0.01, 1, 1, 34, "基本持平，团队新增1人。", []), (COMP_B, FOUNDER_B, 2, 150, 0.0, 0, 0, 22, "营收持平，产品迭代中。", []), (COMP_B, FOUNDER_B, 3, 150, 0.02, 1, 0, 21, "营收微增，新签1家客户。", []), (COMP_B, FOUNDER_B, 4, 147, -0.01, 0, 1, 21, "营收微降，流失1家小客户。", [{"level": "medium", "item": "客户流失", "detail": "流失1家"}]), (COMP_B, FOUNDER_B, 5, 149, 0.03, 1, 0, 20, "营收回升，新签1家。", []), (COMP_B, FOUNDER_B, 6, 145, 0.0, 0, 0, 20, "持平，团队稳定。", []), (COMP_C, FOUNDER_C, 2, 984, 0.18, 3, 0, 62, "强劲增长，半导体客户+3。", []), (COMP_C, FOUNDER_C, 3, 834, 0.15, 2, 0, 58, "持续增长，新增专利1项。", []), (COMP_C, FOUNDER_C, 4, 725, 0.12, 1, 0, 55, "稳定增长，团队扩张。", []), (COMP_C, FOUNDER_C, 5, 647, 0.10, 1, 0, 52, "增速放缓，但趋势向上。", []), (COMP_C, FOUNDER_C, 6, 588, 0.08, 0, 0, 50, "基线月份，稳步起步。", []), (COMP_E, FOUNDER_A, 2, 75, 0.0, 0, 0, 17, "里程碑收入稳定。", []), (COMP_E, FOUNDER_A, 3, 80, 0.07, 0, 0, 18, "小幅增长，靶点推进。", []), (COMP_E, FOUNDER_A, 4, 75, 0.0, 0, 0, 17, "持平，研发正常。", []), (COMP_E, FOUNDER_A, 5, 70, -0.07, 0, 0, 16, "收入微降，无里程碑。", [{"level": "low", "item": "收入波动", "detail": "依赖里程碑付款"}]), (COMP_E, FOUNDER_A, 6, 75, 0.04, 0, 0, 16, "基线月份，稳定。", []), ] for comp_id, founder_id, months_ago, revenue, growth, new_c, churned, hc, summary, concerns in historical_data: m = NOW.month - months_ago y = NOW.year if m <= 0: m += 12 y -= 1 db.add(MonthlyReport( company_id=comp_id, period_year=y, period_month=m, status="reviewed", submitted_by=founder_id, submitted_at=D(months_ago * 30 + 10), reviewed_by=MGR_ID, reviewed_at=D(months_ago * 30 + 5), raw_content=f"本月营收{revenue}万，环比{'增长' if growth > 0 else '持平' if growth == 0 else '下降'}{abs(growth)*100:.0f}%。", structured_data={"revenue": revenue, "mom_growth": growth, "new_customers": new_c, "churned": churned, "headcount": hc}, ai_summary=summary, ai_concerns=concerns if concerns else None, )) await db.flush() print(" ✓ 月报 (5 当月 + 20 历史 = 25 份)") # ─── 健康度评分 ──────────────────────────────────────────────────────── async def _seed_health(db: AsyncSession) -> None: """最新一期健康度评分 — 14 维度完整评分。""" scores = [ # 智链科技 — 良好 HealthScore(company_id=COMP_A, total_score=78.5, financial_score=82, operational_score=75, ai_commercial_score=80, ai_cost_score=85, org_talent_score=72, product_tech_score=80, market_compete_score=78, governance_score=75, financing_score=80, synergy_score=76, ai_model_product_score=82, data_compliance_score=88, team_tech_score=74, customer_success_score=68, trend="up", evidence_json={"financial": "营收增长15%，现金流健康", "ai_cost": "推理成本降低12%", "data_compliance": "通过等保三级", "ai_poc_count": 3, "compliance_issues": 0}, recommendations_json={"action": "关注客户流失率", "owner": MGR_ID, "review_at": D(-30).isoformat()}, fund_type="early_vc", fund_lifecycle="investment", company_stage="b", industry="ai", strategy="growth"), # 云栈数据 — 中等 HealthScore(company_id=COMP_B, total_score=62.0, financial_score=55, operational_score=68, ai_commercial_score=60, ai_cost_score=65, org_talent_score=70, product_tech_score=72, market_compete_score=58, governance_score=65, financing_score=50, synergy_score=64, ai_model_product_score=68, data_compliance_score=85, team_tech_score=72, customer_success_score=60, trend="stable", evidence_json={"financial": "营收持平，需突破", "financing": "Runway 15月，需启动融资", "ai_poc_count": 1, "compliance_issues": 1}, recommendations_json={"action": "加速融资节奏", "owner": LEAD_ID, "review_at": D(-14).isoformat()}, fund_type="early_vc", fund_lifecycle="investment", company_stage="a", industry="ai", strategy="growth"), # 深瞳智能 — 优秀 HealthScore(company_id=COMP_C, total_score=85.5, financial_score=88, operational_score=85, ai_commercial_score=90, ai_cost_score=82, org_talent_score=83, product_tech_score=88, market_compete_score=86, governance_score=82, financing_score=85, synergy_score=80, ai_model_product_score=86, data_compliance_score=90, team_tech_score=85, customer_success_score=82, trend="up", evidence_json={"financial": "营收增长22%", "market": "半导体渗透加深", "ai_poc_count": 5, "compliance_issues": 0}, recommendations_json={"action": "关注客户集中度", "owner": MGR_ID, "review_at": D(-30).isoformat()}, fund_type="early_vc", fund_lifecycle="growth", company_stage="b", industry="ai", strategy="growth"), # 量子芯微 — 早期风险 HealthScore(company_id=COMP_D, total_score=48.0, financial_score=35, operational_score=50, ai_commercial_score=45, ai_cost_score=55, org_talent_score=52, product_tech_score=60, market_compete_score=42, governance_score=55, financing_score=40, synergy_score=45, ai_model_product_score=58, data_compliance_score=65, team_tech_score=62, customer_success_score=38, trend="down", evidence_json={"financial": "无营收，现金紧张", "financing": "Runway 7月", "ai_poc_count": 0, "compliance_issues": 2}, recommendations_json={"action": "紧急启动天使+轮融资", "owner": GP_ID, "review_at": D(-7).isoformat()}, fund_type="angel", fund_lifecycle="investment", company_stage="seed", industry="hardware", strategy="growth"), # 光合生物 — 中等偏上 HealthScore(company_id=COMP_E, total_score=68.5, financial_score=60, operational_score=72, ai_commercial_score=70, ai_cost_score=65, org_talent_score=68, product_tech_score=75, market_compete_score=65, governance_score=70, financing_score=62, synergy_score=66, ai_model_product_score=72, data_compliance_score=80, team_tech_score=70, customer_success_score=65, trend="stable", evidence_json={"financial": "里程碑收入稳定", "product": "3靶点推进中", "ai_poc_count": 2, "compliance_issues": 0}, recommendations_json={"action": "月报提交及时性改善", "owner": MGR_ID, "review_at": D(-30).isoformat()}, fund_type="early_vc", fund_lifecycle="investment", company_stage="a", industry="biotech", strategy="growth"), ] for s in scores: s.calculated_at = D(5) db.add(s) await db.flush() print(" ✓ 健康度评分 (5 企业 x 14 维度)") # ─── 风险事件 ────────────────────────────────────────────────────────── async def _seed_risks(db: AsyncSession) -> None: risks = [ # COMP_A — 智链科技 RiskEvent(company_id=COMP_A, type="operational", severity="medium", status="in_progress", title="客户流失率环比上升", description="本月流失1家年合同客户，流失率从2%升至4%", evidence_json={"metric": "churn_rate", "current": 0.04, "previous": 0.02, "threshold": 0.03}, suggested_action="联系流失客户了解原因，加强客户成功团队", assigned_to=MGR_ID, due_at=D(-7)), RiskEvent(company_id=COMP_A, type="ai_specific", severity="low", status="closed", title="AI模型推理成本超标", description="上月推理成本占营收比12%，超10%阈值", evidence_json={"metric": "ai_cost_ratio", "current": 0.12, "threshold": 0.10}, suggested_action="优化模型量化策略", assigned_to=MGR_ID, closed_at=D(-5)), RiskEvent(company_id=COMP_A, type="market", severity="medium", status="open", title="竞品低价策略冲击", description="杉数科技以低于我方30%的价格争夺制造业客户", evidence_json={"competitor": "杉数科技", "price_diff": -0.30, "affected_customers": 3}, suggested_action="强化差异化话术，加速多模态产品迭代", assigned_to=LEAD_ID), RiskEvent(company_id=COMP_A, type="financial", severity="low", status="resolved", title="应收账款周期延长", description="2家客户付款周期从30天延至45天", evidence_json={"metric": "ar_days", "current": 45, "previous": 30, "threshold": 60}, suggested_action="与客户协商缩短账期", assigned_to=MGR_ID, closed_at=D(-15)), # COMP_B — 云栈数据 RiskEvent(company_id=COMP_B, type="financial", severity="high", status="assigned", title="营收增长停滞", description="连续2个月环比0%增长，未达预期", evidence_json={"metric": "revenue_growth", "current": 0.0, "expected": 0.15, "months": 2}, suggested_action="与创始人讨论增长策略，评估产品定价和市场拓展", assigned_to=LEAD_ID, due_at=D(-3)), RiskEvent(company_id=COMP_B, type="financial", severity="high", status="open", title="现金Runway缩短至15个月", description="按当前消耗速度，现金仅可维持15个月", evidence_json={"metric": "runway_months", "current": 15, "threshold": 18}, suggested_action="启动A+轮融资准备，控制招聘节奏", assigned_to=GP_ID), RiskEvent(company_id=COMP_B, type="market", severity="medium", status="open", title="金融行业准入门槛高", description="银行客户POC周期长达6-12月，转化率低", evidence_json={"metric": "poc_conversion", "current": 0.15, "expected": 0.30}, suggested_action="调整目标客户画像，探索保险/证券赛道", assigned_to=MGR_ID), # COMP_C — 深瞳智能 RiskEvent(company_id=COMP_C, type="org", severity="medium", status="open", title="核心技术人员稳定性下降", description="CTO提及外部机会，稳定性评分从0.85降至0.65", evidence_json={"metric": "cto_stability", "current": 0.65, "previous": 0.85}, suggested_action="了解CTO诉求，评估股权激励调整", assigned_to=LEAD_ID), RiskEvent(company_id=COMP_C, type="market", severity="medium", status="in_progress", title="客户集中度过高", description="半导体行业占比60%，周期下行风险大", evidence_json={"metric": "customer_concentration", "current": 0.60, "threshold": 0.50, "industry": "semiconductor"}, suggested_action="拓展新能源、消费电子等新行业客户", assigned_to=MGR_ID, due_at=D(-30)), RiskEvent(company_id=COMP_C, type="operational", severity="low", status="closed", title="产品交付延迟", description="1条产线视觉系统部署延迟2周", evidence_json={"metric": "delivery_delay_days", "current": 14, "threshold": 7}, suggested_action="增加现场实施人员", assigned_to=MGR_ID, closed_at=D(-20)), # COMP_D — 量子芯微 RiskEvent(company_id=COMP_D, type="financial", severity="critical", status="open", title="现金Runway不足7个月", description="按当前消耗速度，现金仅可维持7个月", evidence_json={"metric": "runway_months", "current": 7, "threshold": 9}, suggested_action="紧急启动天使+轮融资，准备BP和财务预测", assigned_to=GP_ID, due_at=D(-14)), RiskEvent(company_id=COMP_D, type="technical", severity="high", status="in_progress", title="NPU流片进度风险", description="EDA工具链兼容性问题可能导致流片延迟1-2月", evidence_json={"metric": "tapeout_delay_risk", "risk_level": "high", "eta_impact": "1-2月"}, suggested_action="协调备用EDA工具，增加验证人力", assigned_to=GP_ID, due_at=D(-7)), RiskEvent(company_id=COMP_D, type="market", severity="medium", status="open", title="边缘NPU赛道巨头入场", description="英伟达发布边缘端Jetson新品，可能挤压创业公司空间", evidence_json={"competitor": "NVIDIA", "product": "Jetson Orin Nano", "threat_level": "medium"}, suggested_action="强化低功耗差异化，加速客户绑定", assigned_to=LEAD_ID), # COMP_E — 光合生物 RiskEvent(company_id=COMP_E, type="operational", severity="medium", status="resolved", title="月报延迟提交", description="上月月报延迟20天提交", evidence_json={"metric": "report_delay_days", "current": 20, "threshold": 5}, suggested_action="与创始人沟通提交规范，设置自动提醒", assigned_to=MGR_ID, due_at=D(-10), closed_at=D(-3)), RiskEvent(company_id=COMP_E, type="financial", severity="medium", status="open", title="研发投入产出周期长", description="3靶点尚处先导化合物阶段，距临床还需12-18月", evidence_json={"metric": "time_to_clinic", "current": 18, "threshold": 12, "targets": 3}, suggested_action="评估是否引入大药企联合开发", assigned_to=LEAD_ID), ] for r in risks: r.identified_at = D(15) db.add(r) await db.flush() print(" ✓ 风险事件 (15 条)") # ─── 弱信号 ──────────────────────────────────────────────────────────── async def _seed_weak_signals(db: AsyncSession) -> None: signals = [ WeakSignal(company_id=COMP_B, signal_type="market", source="行业论坛", content="竞品发布类似数据治理产品，定价低30%", confidence=0.72, risk_probability=0.65, status="alerted", correlation_result={"related_signals": 2, "pattern": "竞品压力增加"}), WeakSignal(company_id=COMP_C, signal_type="org", source="LinkedIn", content="CTO 更新简历，标注'开放看机会'", confidence=0.85, risk_probability=0.78, status="alerted", correlation_result={"related_signals": 1, "pattern": "核心人员流失风险"}), WeakSignal(company_id=COMP_D, signal_type="technical", source="GitHub", content="核心开源依赖库停止维护，需迁移", confidence=0.68, risk_probability=0.45, status="correlated"), WeakSignal(company_id=COMP_A, signal_type="sentiment", source="客户反馈", content="大客户提及预算削减计划", confidence=0.60, risk_probability=0.55, status="new"), WeakSignal(company_id=COMP_E, signal_type="market", source="FDA新闻", content="同类靶点药物获FDA快速审批", confidence=0.75, risk_probability=0.30, status="new"), WeakSignal(company_id=COMP_C, signal_type="market", source="半导体行业协会", content="半导体周期指标连续2月下行", confidence=0.70, risk_probability=0.62, status="correlated", correlation_result={"related_signals": 1, "pattern": "行业周期风险"}), WeakSignal(company_id=COMP_A, signal_type="technical", source="GitHub", content="竞品开源多模态决策引擎，star数暴涨", confidence=0.65, risk_probability=0.40, status="new"), WeakSignal(company_id=COMP_B, signal_type="org", source="脉脉", content="云栈数据销售总监更新简历", confidence=0.78, risk_probability=0.60, status="alerted", correlation_result={"related_signals": 1, "pattern": "销售负责人不稳定"}), WeakSignal(company_id=COMP_D, signal_type="market", source="英伟达发布会", content="英伟达发布边缘端Jetson Orin Nano，价格下探", confidence=0.82, risk_probability=0.70, status="alerted", correlation_result={"related_signals": 2, "pattern": "巨头入场边缘AI"}), ] for s in signals: s.detected_at = D(7) db.add(s) await db.flush() print(" ✓ 弱信号 (9 条)") # ─── 任务 ────────────────────────────────────────────────────────────── async def _seed_tasks(db: AsyncSession) -> None: tasks = [ Task(tenant_id=TENANT_ID, company_id=COMP_A, title="联系流失客户进行回访", description="了解客户流失原因，制定挽留方案", status="in_progress", priority="high", assigned_to=MGR_ID, source_type="risk", due_at=D(-3)), Task(tenant_id=TENANT_ID, company_id=COMP_B, title="与云栈数据创始人讨论增长策略", description="评估产品定价和市场拓展方案", status="todo", priority="high", assigned_to=LEAD_ID, source_type="risk", due_at=D(-1)), Task(tenant_id=TENANT_ID, company_id=COMP_D, title="准备天使+轮BP", description="包含财务预测和产品路线图", status="todo", priority="urgent", assigned_to=GP_ID, source_type="risk", due_at=D(-14)), Task(tenant_id=TENANT_ID, company_id=COMP_C, title="评估CTO股权激励调整方案", description="了解CTO诉求，设计retention package", status="todo", priority="medium", assigned_to=LEAD_ID, source_type="manual", due_at=D(-7)), Task(tenant_id=TENANT_ID, company_id=COMP_A, title="审阅智链科技本月月报", description="关注AI成本降低和客户流失", status="done", priority="medium", assigned_to=MGR_ID, source_type="report", completed_at=D(-3)), Task(tenant_id=TENANT_ID, company_id=COMP_E, title="跟进光合生物月报提交及时性", description="设置自动提醒，沟通提交规范", status="done", priority="medium", assigned_to=MGR_ID, source_type="manual", completed_at=D(-2)), ] for t in tasks: db.add(t) await db.flush() print(" ✓ 任务 (6 条)") # ─── OKR ─────────────────────────────────────────────────────────────── async def _seed_okrs(db: AsyncSession) -> None: okrs = [ OKR(company_id=COMP_A, quarter=f"{NOW.year}-Q{(NOW.month - 1) // 3 + 1}", objective="实现季度营收3000万，客户净增8家", key_results=[{"kr": "季度营收达3000万", "current": 2400, "target": 3000, "progress": 0.80}, {"kr": "新签客户8家", "current": 5, "target": 8, "progress": 0.625}, {"kr": "客户流失率<3%", "current": 0.04, "target": 0.03, "progress": 0.75}], alignment_score=82.0, status="active", deviation_alerts=[{"kr": "客户流失率", "current": "4%", "target": "<3%", "severity": "medium"}]), OKR(company_id=COMP_B, quarter=f"{NOW.year}-Q{(NOW.month - 1) // 3 + 1}", objective="完成A+轮融资，营收突破500万/月", key_results=[{"kr": "月营收达500万", "current": 150, "target": 500, "progress": 0.30}, {"kr": "完成A+轮融资", "current": 0, "target": 1, "progress": 0.0}, {"kr": "签约3家金融客户", "current": 1, "target": 3, "progress": 0.33}], alignment_score=55.0, status="active", deviation_alerts=[{"kr": "月营收", "current": "150万", "target": "500万", "severity": "high"}, {"kr": "融资进度", "current": "未启动", "target": "完成", "severity": "high"}]), OKR(company_id=COMP_C, quarter=f"{NOW.year}-Q{(NOW.month - 1) // 3 + 1}", objective="半导体行业营收占比70%，新增专利5项", key_results=[{"kr": "半导体营收占比70%", "current": 0.60, "target": 0.70, "progress": 0.857}, {"kr": "新增专利5项", "current": 2, "target": 5, "progress": 0.40}], alignment_score=88.0, status="active", deviation_alerts=[]), ] for o in okrs: db.add(o) await db.flush() print(" ✓ OKR (3 条)") # ─── 里程碑 ──────────────────────────────────────────────────────────── async def _seed_milestones(db: AsyncSession) -> None: # 智链科技里程碑树 m1 = MilestoneTree(company_id=COMP_A, name="B轮融资", is_current=True, status="in_progress", target_date=D(-90), description="B轮1亿人民币，估值8亿") db.add(m1); await db.flush() m2 = MilestoneTree(company_id=COMP_A, name="产品v3.0发布", parent_id=m1.id, is_current=True, status="completed", actual_date=D(-20), description="多模态供应链决策引擎") m3 = MilestoneTree(company_id=COMP_A, name="客户数突破50", parent_id=m1.id, is_current=True, status="in_progress", target_date=D(-60)) m4 = MilestoneTree(company_id=COMP_A, name="海外市场拓展", parent_id=m1.id, is_current=False, status="planned", target_date=D(-180), description="东南亚试点 — 备选路径", ai_analysis={"risk": "海外合规成本高", "opportunity": "市场先发优势"}) for m in [m2, m3, m4]: db.add(m) # 云栈数据里程碑 m5 = MilestoneTree(company_id=COMP_B, name="A+轮融资", is_current=True, status="in_progress", target_date=D(-45), description="A+轮3000万人民币") db.add(m5); await db.flush() m6 = MilestoneTree(company_id=COMP_B, name="数据治理平台v2.0", parent_id=m5.id, is_current=True, status="completed", actual_date=D(-10)) m7 = MilestoneTree(company_id=COMP_B, name="金融行业突破", parent_id=m5.id, is_current=True, status="in_progress", target_date=D(-30)) for m in [m6, m7]: db.add(m) await db.flush() print(" ✓ 里程碑 (7 节点)") # ─── 协同机会 ────────────────────────────────────────────────────────── async def _seed_synergies(db: AsyncSession) -> None: synergies = [ SynergyOpportunity(tenant_id=TENANT_ID, type="customer", company_a_id=COMP_A, company_b_id=COMP_C, title="智链科技 → 深瞳智能：制造业客户共享", description="智链科技的3家制造业客户有质检需求，可引荐深瞳智能。", match_reason="AI分析：客户行业重叠度高，需求互补", status="confirmed", authorized=True), SynergyOpportunity(tenant_id=TENANT_ID, type="tech", company_a_id=COMP_B, company_b_id=COMP_D, title="云栈数据 → 量子芯微：数据平台支持芯片设计仿真", description="云栈数据平台可处理量子芯微的仿真数据集。", match_reason="技术栈匹配，数据平台可降低芯片仿真IT成本40%", status="discovered", authorized=False), SynergyOpportunity(tenant_id=TENANT_ID, type="talent", company_a_id=COMP_C, company_b_id=COMP_E, title="深瞳智能 → 光合生物：AI算法人才推荐", description="深瞳智能有CV算法人才储备，光合生物需AI药物筛选算法工程师。", match_reason="技能匹配度85%，工作地点同城", status="executing", authorized=True, effect_result={"candidates_referred": 3, "interviewed": 2, "offered": 1}), ] for s in synergies: db.add(s) await db.flush() print(" ✓ 协同机会 (3 条)") # ─── 决策前哨 ────────────────────────────────────────────────────────── async def _seed_sentinels(db: AsyncSession) -> None: sentinels = [ DecisionSentinel(company_id=COMP_B, decision_type="funding", title="A+轮融资时机选择：现在 vs 等待产品验证", description="营收增长停滞，但产品v2.0刚发布。是否现在融资还是等Q2数据验证后融资？", signals=[{"source": "月报", "signal": "营收停滞2月"}, {"source": "市场", "signal": "竞品定价低30%"}], scenarios={"A": {"name": "现在融资", "pros": "现金充裕", "cons": "估值可能偏低", "probability": 0.6}, "B": {"name": "等待Q2验证", "pros": "估值更高", "cons": "现金Runway风险", "probability": 0.4}}, status="analyzed"), DecisionSentinel(company_id=COMP_C, decision_type="hiring", title="CTO retention：股权激励 vs 替换", description="CTO稳定性下降，需决定加大股权激励还是提前寻找替代者。", signals=[{"source": "LinkedIn", "signal": "CTO更新简历"}, {"source": "弱信号", "signal": "稳定性评分0.65"}], scenarios={"A": {"name": "加大股权激励", "pros": "保留技术连续性", "cons": "股权稀释", "probability": 0.7}, "B": {"name": "启动替换搜索", "pros": "降低单点风险", "cons": "过渡期风险", "probability": 0.3}}, status="identified"), DecisionSentinel(company_id=COMP_D, decision_type="funding", title="天使+轮 vs Pre-A：融资策略选择", description="现金Runway 7月，需快速融资。天使+轮快但金额小，Pre-A慢但可支撑更长。", signals=[{"source": "财务", "signal": "Runway 7月"}, {"source": "产品", "signal": "NPU原型流片在即"}], scenarios={"A": {"name": "天使+轮 1500万", "pros": "3个月内close", "cons": "只够6月", "probability": 0.55}, "B": {"name": "Pre-A 5000万", "pros": "支撑18月", "cons": "需6月", "probability": 0.45}}, status="acted"), ] for s in sentinels: s.identified_at = D(10) db.add(s) await db.flush() print(" ✓ 决策前哨 (3 条)") # ─── 行为助推 ────────────────────────────────────────────────────────── async def _seed_nudges(db: AsyncSession) -> None: nudges = [ NudgeRecord(company_id=COMP_A, nudge_type="anchoring", context="月报审阅 — 客户流失率", message="同类AI供应链企业平均客户流失率为2.5%，智链科技本月4%高于基准。", target_user_id=MGR_ID, accepted=True, effect_result={"action": "联系流失客户", "result": "挽留1家"}), NudgeRecord(company_id=COMP_B, nudge_type="loss_aversion", context="融资时机提醒", message="如果不启动A+轮融资，预计6个月后现金耗尽，错失产品验证窗口。", target_user_id=LEAD_ID, accepted=True, effect_result={"action": "约见创始人讨论融资", "result": "启动BP准备"}), NudgeRecord(company_id=COMP_D, nudge_type="social_proof", context="融资策略", message="Portfolio中3家类似阶段企业选择了Pre-A路线，平均估值高于天使+轮2倍。", target_user_id=GP_ID, accepted=False, effect_result={"action": "选择天使+轮", "result": "快速close"}), NudgeRecord(company_id=COMP_C, nudge_type="timing", context="CTO稳定性", message="融资前是关键人才挽留的最佳时机，建议在B+轮前完成CTO股权激励调整。", target_user_id=LEAD_ID, accepted=True, effect_result={"action": "设计retention package", "result": "进行中"}), ] for n in nudges: db.add(n) await db.flush() print(" ✓ 行为助推 (4 条)") # ─── 重大事项 ────────────────────────────────────────────────────────── async def _seed_events(db: AsyncSession) -> None: events = [ MajorEvent(company_id=COMP_A, event_type="product", title="产品v3.0多模态引擎发布", severity="high", description="支持文本+图像+时序数据的多模态供应链决策引擎上线", source="monthly_report", status="addressed", occurred_at=D(20)), MajorEvent(company_id=COMP_C, event_type="personnel", title="CTO更新简历", severity="high", description="CTO在LinkedIn更新简历，标注'开放看机会'", source="weak_signal", status="confirmed", occurred_at=D(7)), MajorEvent(company_id=COMP_D, event_type="funding", title="天使+轮融资启动", severity="critical", description="现金Runway不足7月，启动天使+轮融资", source="manual", status="addressed", occurred_at=D(3)), MajorEvent(company_id=COMP_E, event_type="product", title="3个靶点进入先导化合物阶段", severity="medium", description="AI药物筛选管线推进，3个靶点通过虚拟筛选进入实验验证", source="monthly_report", status="confirmed", occurred_at=D(15)), ] for e in events: db.add(e) await db.flush() print(" ✓ 重大事项 (4 条)") # ─── 投资协议 ────────────────────────────────────────────────────────── async def _seed_agreements(db: AsyncSession) -> None: agreements = [ InvestmentAgreement(company_id=COMP_A, title="B轮投资协议", signed_at=datetime(2023, 6, 1, tzinfo=UTC), key_clauses={"valuation": "8亿人民币", "investment": "1亿", "board_seat": "1席", "anti_dilution": "加权平均", "liquidation_pref": "1x非参与"}, monitoring_rules=[{"clause": "反稀释", "trigger": "下一轮估值低于8亿", "action": "通知GP"}], status="active"), InvestmentAgreement(company_id=COMP_B, title="A轮投资协议", signed_at=datetime(2023, 3, 1, tzinfo=UTC), key_clauses={"valuation": "3亿", "investment": "5000万", "board_seat": "1席", "redemption_right": "5年后可要求回购"}, monitoring_rules=[{"clause": "回购权", "trigger": "2028-03-01", "action": "提前6月提醒"}], status="active"), InvestmentAgreement(company_id=COMP_C, title="B轮投资协议", signed_at=datetime(2022, 12, 1, tzinfo=UTC), key_clauses={"valuation": "10亿", "investment": "2亿", "board_seat": "1席", "info_rights": "月报+季度财务"}, monitoring_rules=[{"clause": "信息权", "trigger": "月报延迟>5天", "action": "通知投后负责人"}], status="active"), ] for a in agreements: db.add(a) await db.flush() print(" ✓ 投资协议 (3 份)") # ─── 董事会 ──────────────────────────────────────────────────────────── async def _seed_boards(db: AsyncSession) -> None: boards = [ BoardMeeting(company_id=COMP_A, title="2025年Q2董事会", meeting_at=D(-7), status="scheduled", agenda=[{"item": "Q2经营汇报", "duration": "30min"}, {"item": "B轮融资进展", "duration": "20min"}, {"item": "客户流失分析", "duration": "15min"}], materials_summary="AI摘要：Q2营收2400万(完成80%)，客户净增5家但流失率上升，B轮估值预期8-10亿。", questions=[{"q": "客户流失率上升的原因是什么？", "priority": "high"}, {"q": "B轮预计何时close？", "priority": "medium"}]), BoardMeeting(company_id=COMP_C, title="2025年Q2董事会", meeting_at=D(-14), status="completed", agenda=[{"item": "Q2经营汇报", "duration": "30min"}, {"item": "半导体行业拓展", "duration": "20min"}], minutes="Q2营收3600万超预期，半导体占比60%。专利新增2项。CTO稳定性需关注。", resolutions=[{"item": "批准CTO股权激励调整方案", "status": "执行中"}, {"item": "Q3启动半导体行业专项销售", "status": "已启动"}]), ] for b in boards: db.add(b) await db.flush() print(" ✓ 董事会 (2 场)") # ─── BML 假设 ────────────────────────────────────────────────────────── async def _seed_hypotheses(db: AsyncSession) -> None: hypotheses = [ Hypothesis(company_id=COMP_A, hypothesis="降低定价10%可提升客户签约率30%", experiment="对3家潜在客户提供10%折扣试用", data={"customers_tested": 3, "signed": 2, "conversion_rate": 0.67}, conclusion="部分验证：签约率提升67%（超预期30%），但需观察折扣结束后续约率", status="validated"), Hypothesis(company_id=COMP_B, hypothesis="金融行业客户愿意为数据合规功能支付20%溢价", experiment="向2家金融客户展示合规模块并报价溢价20%", data={"customers_tested": 2, "interested": 1, "signed": 0}, conclusion="待验证：1家表示感兴趣但需内部审批，尚未签约", status="measuring"), Hypothesis(company_id=COMP_C, hypothesis="半导体客户对边缘端质检需求年增长50%", experiment="调研5家半导体客户未来12月质检预算", data={"surveyed": 3, "budget_increase_avg": 0.35}, conclusion="未达预期：平均预算增长35%低于50%假设，但基础需求强劲", status="learning"), ] for h in hypotheses: db.add(h) await db.flush() print(" ✓ BML 假设 (3 条)") # ─── AAR 复盘 ────────────────────────────────────────────────────────── async def _seed_aars(db: AsyncSession) -> None: aars = [ AARRecord(company_id=COMP_A, trigger_event="Q1客户流失率升至4%", original_plan="Q1客户流失率控制在2%以内", actual_result="Q1流失率4%，流失2家年合同客户", gap_analysis="客户成功团队响应延迟，流失客户反馈'技术支持不够及时'。竞品以低价策略抢客。", lessons={"what_expected": "流失率2%", "what_happened": "4%", "why_gap": "客户成功人手不足+竞品低价", "what_learned": "需主动客户健康度监控", "what_to_change": "增加客户成功HC，建立预警机制"}, improvements=[{"action": "增加2名客户成功经理", "status": "已完成", "owner": "创始人"}, {"action": "建立客户健康度周报", "status": "进行中", "owner": "投资经理"}]), AARRecord(company_id=COMP_C, trigger_event="CTO稳定性下降", original_plan="核心技术团队稳定，无关键人员流失", actual_result="CTO更新简历，稳定性评分0.65", gap_analysis="CTO反馈股权激励不足，技术路线存在分歧。", lessons={"what_expected": "CTO长期稳定", "what_happened": "稳定性下降", "why_gap": "股权激励未随估值增长调整", "what_learned": "关键人才激励需动态调整", "what_to_change": "建立定期激励review机制"}, improvements=[{"action": "CTO股权激励调整方案", "status": "设计中", "owner": "投后负责人"}]), ] for a in aars: db.add(a) await db.flush() print(" ✓ AAR 复盘 (2 条)") # ─── 团队成员 ────────────────────────────────────────────────────────── async def _seed_team_members(db: AsyncSession) -> None: members = [ TeamMember(company_id=COMP_A, name="陈智链", role="CEO/创始人", is_key_person=True, joined_at=datetime(2021, 3, 1, tzinfo=UTC), stability_score=0.92), TeamMember(company_id=COMP_A, name="张技术", role="CTO", is_key_person=True, joined_at=datetime(2021, 4, 1, tzinfo=UTC), stability_score=0.85), TeamMember(company_id=COMP_A, name="李销售", role="VP销售", is_key_person=True, joined_at=datetime(2022, 1, 1, tzinfo=UTC), stability_score=0.80), TeamMember(company_id=COMP_C, name="赵深瞳", role="CEO/创始人", is_key_person=True, joined_at=datetime(2020, 6, 1, tzinfo=UTC), stability_score=0.90), TeamMember(company_id=COMP_C, name="孙算法", role="CTO", is_key_person=True, joined_at=datetime(2020, 7, 1, tzinfo=UTC), stability_score=0.65), TeamMember(company_id=COMP_C, name="周产品", role="产品总监", is_key_person=False, joined_at=datetime(2021, 3, 1, tzinfo=UTC), stability_score=0.78), TeamMember(company_id=COMP_B, name="刘云栈", role="CEO/创始人", is_key_person=True, joined_at=datetime(2022, 1, 1, tzinfo=UTC), stability_score=0.88), ] for m in members: db.add(m) await db.flush() print(" ✓ 团队成员 (7 人)") # ─── 人才画像 ────────────────────────────────────────────────────────── async def _seed_talents(db: AsyncSession) -> None: talents = [ TalentProfile(tenant_id=TENANT_ID, name="孙算法", current_role="CTO", current_company="深瞳智能", skills=["计算机视觉", "PyTorch", "团队管理", "边缘计算"], experience_years=12, performance_rating=4.5, potential_rating=4.0, nine_box="star", flow_prediction={"probability": 0.65, "timeline": "3-6月", "triggers": ["股权激励不足", "技术路线分歧"]}, status="active"), TalentProfile(tenant_id=TENANT_ID, name="李销售", current_role="VP销售", current_company="智链科技", skills=["B2B销售", "制造业客户", "团队管理"], experience_years=10, performance_rating=4.2, potential_rating=3.8, nine_box="high_potential", flow_prediction={"probability": 0.20, "timeline": "12月+", "triggers": []}, status="active"), TalentProfile(tenant_id=TENANT_ID, name="吴架构", current_role="首席架构师", current_company="云栈数据", skills=["分布式系统", "数据治理", "Kubernetes"], experience_years=15, performance_rating=4.8, potential_rating=4.5, nine_box="star", flow_prediction={"probability": 0.15, "timeline": "稳定", "triggers": []}, status="active"), ] for t in talents: db.add(t) await db.flush() print(" ✓ 人才画像 (3 人)") # ─── 审计日志 ────────────────────────────────────────────────────────── async def _seed_audit_logs(db: AsyncSession) -> None: logs = [ AuditLog(tenant_id=TENANT_ID, user_id=MGR_ID, action="login", resource_type="auth", ip="192.168.1.100", created_at=D(0)), AuditLog(tenant_id=TENANT_ID, user_id=MGR_ID, action="view", resource_type="company", resource_id=COMP_A, ip="192.168.1.100", created_at=D(0)), AuditLog(tenant_id=TENANT_ID, user_id=MGR_ID, action="view", resource_type="report", resource_id="report_001", ip="192.168.1.100", created_at=D(0)), AuditLog(tenant_id=TENANT_ID, user_id=LEAD_ID, action="view", resource_type="risk", resource_id="risk_001", ip="192.168.1.101", created_at=D(1)), AuditLog(tenant_id=TENANT_ID, user_id=LEAD_ID, action="update", resource_type="risk", resource_id="risk_001", detail_json={"field": "status", "old": "open", "new": "assigned"}, ip="192.168.1.101", created_at=D(1)), AuditLog(tenant_id=TENANT_ID, user_id=GP_ID, action="view", resource_type="dashboard", ip="192.168.1.102", created_at=D(2)), AuditLog(tenant_id=TENANT_ID, user_id=GP_ID, action="ai_call", resource_type="weekly_brief", detail_json={"tokens": 3500, "model": "qwen-max"}, ip="192.168.1.102", created_at=D(2)), AuditLog(tenant_id=TENANT_ID, user_id=ADMIN_ID, action="view", resource_type="audit_logs", ip="192.168.1.103", created_at=D(3)), AuditLog(tenant_id=TENANT_ID, user_id=FOUNDER_A, action="create", resource_type="monthly_report", detail_json={"company": "智链科技", "period": f"{NOW.year}-{NOW.month-1}"}, ip="10.0.0.50", created_at=D(10)), AuditLog(tenant_id=TENANT_ID, user_id=FOUNDER_B, action="create", resource_type="monthly_report", detail_json={"company": "云栈数据", "period": f"{NOW.year}-{NOW.month-1}"}, ip="10.0.0.51", created_at=D(8)), ] for l in logs: db.add(l) await db.flush() print(" ✓ 审计日志 (10 条)") # ─── 健康度历史趋势（近 6 个月） ────────────────────────────────────── async def _seed_health_history(db: AsyncSession) -> None: """每家企业近 6 个月健康度趋势，用于趋势图展示。""" # 基线分数和月度变化趋势 trends = [ # (company_id, base_score, monthly_deltas) (COMP_A, 70.0, [+1.5, +2.0, +1.5, +2.0, +1.5, +0.0]), # 上升趋势 (COMP_B, 65.0, [-0.5, -0.5, +0.0, -0.5, -1.0, +0.0]), # 下降趋势 (COMP_C, 75.0, [+1.5, +2.0, +2.5, +2.0, +1.5, +1.0]), # 强上升 (COMP_D, 58.0, [-2.0, -2.0, -2.5, -1.5, -1.0, -1.0]), # 持续下降 (COMP_E, 65.0, [+0.5, +1.0, +0.5, +1.0, +0.5, +0.0]), # 缓慢上升 ] for comp_id, base, deltas in trends: score = base for i, delta in enumerate(deltas): score += delta m = (NOW.month - i - 1) if (NOW.month - i - 1) > 0 else (12 + NOW.month - i - 1) y = NOW.year if (NOW.month - i - 1) > 0 else NOW.year - 1 db.add(HealthScore( company_id=comp_id, total_score=round(score, 1), financial_score=round(score * 1.05, 1), operational_score=round(score * 0.95, 1), ai_commercial_score=round(score * 1.02, 1), ai_cost_score=round(score * 1.08, 1), org_talent_score=round(score * 0.92, 1), product_tech_score=round(score * 1.03, 1), market_compete_score=round(score * 0.98, 1), governance_score=round(score * 0.96, 1), financing_score=round(score * 0.94, 1), synergy_score=round(score * 0.97, 1), ai_model_product_score=round(score * 1.04, 1), data_compliance_score=round(score * 1.12, 1), team_tech_score=round(score * 0.95, 1), customer_success_score=round(score * 0.87, 1), trend="up" if delta > 0 else "down" if delta < 0 else "stable", evidence_json={"historical": True, "month": f"{y}-{m:02d}"}, calculated_at=datetime(y, m, 1, tzinfo=UTC), fund_type="early_vc", fund_lifecycle="investment", company_stage="b", industry="ai", strategy="growth", )) await db.flush() print(" ✓ 健康度历史趋势 (5 企业 x 6 月)") # ─── 基金 ────────────────────────────────────────────────────────────── async def _seed_funds(db: AsyncSession) -> None: """基金信息 + 企业-基金关联。""" fund1 = Fund(tenant_id=TENANT_ID, name="远见三期人民币基金", fund_type="early_vc", strategy="growth", established_date=datetime(2023, 1, 1).date(), total_lifespan_months=84, investment_period_months=48, lp_composition_json={"government": 30, "market": 50, "corporate": 20}, primary_market="china_mainland") fund2 = Fund(tenant_id=TENANT_ID, name="远见天使基金", fund_type="angel", strategy="growth", established_date=datetime(2024, 1, 1).date(), total_lifespan_months=72, investment_period_months=36, lp_composition_json={"government": 20, "market": 60, "corporate": 20}, primary_market="china_mainland") for f in [fund1, fund2]: db.add(f) await db.flush() links = [ CompanyFundLink(company_id=COMP_A, fund_id=fund1.id, investment_date=datetime(2023, 6, 1).date(), investment_stage="b", round="B轮", amount=5000, ownership_pct=12.5, is_current=True), CompanyFundLink(company_id=COMP_B, fund_id=fund1.id, investment_date=datetime(2023, 3, 1).date(), investment_stage="a", round="A轮", amount=2000, ownership_pct=15.0, is_current=True), CompanyFundLink(company_id=COMP_C, fund_id=fund1.id, investment_date=datetime(2022, 12, 1).date(), investment_stage="b", round="B轮", amount=8000, ownership_pct=20.0, is_current=True), CompanyFundLink(company_id=COMP_E, fund_id=fund1.id, investment_date=datetime(2023, 9, 1).date(), investment_stage="a", round="A轮", amount=3000, ownership_pct=10.0, is_current=True), CompanyFundLink(company_id=COMP_D, fund_id=fund2.id, investment_date=datetime(2024, 3, 1).date(), investment_stage="seed", round="天使轮", amount=500, ownership_pct=8.0, is_current=True), ] for l in links: db.add(l) await db.flush() print(" ✓ 基金 (2 支) + 关联 (5 条)") # ─── 多主体画像 ──────────────────────────────────────────────────────── async def _seed_profiles(db: AsyncSession) -> None: """投资机构、基金、投资经理画像。""" firm = FirmProfile(tenant_id=TENANT_ID, name="远见资本", focus_areas={"sectors": ["AI", "硬科技", "生物医药"], "stages": ["seed", "A", "B"]}, stage_preference="seed-B", description="专注早期科技投资，管理规模15亿人民币。") db.add(firm) await db.flush() db.add(FundProfile(firm_id=firm.id, name="远见三期人民币基金", fund_size="5亿人民币", vintage_year=2023, strategy="早期科技，AI+硬科技为主，赋能式投后管理。")) db.add(FundProfile(firm_id=firm.id, name="远见天使基金", fund_size="1亿人民币", vintage_year=2024, strategy="天使阶段，聚焦AI应用和芯片设计。")) db.add(ManagerProfile(firm_id=firm.id, user_id=MGR_ID, name="王经理", focus_areas={"sectors": ["AI/供应链", "AI/数据平台"], "companies": [COMP_A, COMP_B]}, portfolio_count=2)) db.add(ManagerProfile(firm_id=firm.id, user_id=LEAD_ID, name="李投后", focus_areas={"sectors": ["AI/计算机视觉", "AI/医疗"], "companies": [COMP_C, COMP_E]}, portfolio_count=2)) await db.flush() print(" ✓ 画像 (机构1 + 基金2 + 经理2)") # ─── 评价模板 ────────────────────────────────────────────────────────── async def _seed_evaluation_templates(db: AsyncSession) -> None: """6 轴动态评价模板 — 覆盖常见组合。""" templates = [ EvaluationTemplate(tenant_id=TENANT_ID, name="早期VC-AI-B轮-成长策略", fund_type="early_vc", fund_lifecycle="investment", company_stage="b", industry="ai", strategy="growth", investor_type="investor", weights_json={"financial": 0.12, "operational": 0.10, "ai_commercial": 0.12, "ai_cost": 0.08, "org_talent": 0.10, "product_tech": 0.12, "market_compete": 0.10, "governance": 0.06, "financing": 0.08, "synergy": 0.04, "ai_model_product": 0.04, "data_compliance": 0.02, "team_tech": 0.01, "customer_success": 0.01}, enabled_dimensions=["financial", "operational", "ai_commercial", "ai_cost", "org_talent", "product_tech", "market_compete", "governance", "financing", "synergy", "ai_model_product", "data_compliance", "team_tech", "customer_success"], disabled_dimensions=[], custom_metrics_json={"ai_poc_count": {"label": "AI PoC 数量", "source": "月报"}, "compliance_issues": {"label": "合规整改项", "source": "审计"}}, is_default=True), EvaluationTemplate(tenant_id=TENANT_ID, name="天使-硬件-种子期-成长策略", fund_type="angel", fund_lifecycle="investment", company_stage="seed", industry="hardware", strategy="growth", investor_type="gp", weights_json={"financial": 0.06, "operational": 0.08, "ai_commercial": 0.08, "ai_cost": 0.06, "org_talent": 0.15, "product_tech": 0.20, "market_compete": 0.10, "governance": 0.05, "financing": 0.12, "synergy": 0.03, "ai_model_product": 0.04, "data_compliance": 0.01, "team_tech": 0.02, "customer_success": 0.00}, enabled_dimensions=["financial", "operational", "ai_commercial", "ai_cost", "org_talent", "product_tech", "market_compete", "governance", "financing", "synergy", "ai_model_product", "data_compliance", "team_tech"], disabled_dimensions=["customer_success"], is_default=True), EvaluationTemplate(tenant_id=TENANT_ID, name="早期VC-生物医疗-A轮-成长策略", fund_type="early_vc", fund_lifecycle="investment", company_stage="a", industry="biotech", strategy="growth", investor_type="investor", weights_json={"financial": 0.08, "operational": 0.10, "ai_commercial": 0.10, "ai_cost": 0.06, "org_talent": 0.12, "product_tech": 0.18, "market_compete": 0.08, "governance": 0.08, "financing": 0.10, "synergy": 0.03, "ai_model_product": 0.04, "data_compliance": 0.03, "team_tech": 0.00, "customer_success": 0.00}, enabled_dimensions=["financial", "operational", "ai_commercial", "ai_cost", "org_talent", "product_tech", "market_compete", "governance", "financing", "synergy", "ai_model_product", "data_compliance"], disabled_dimensions=["team_tech", "customer_success"], is_default=True), ] for t in templates: db.add(t) await db.flush() print(" ✓ 评价模板 (3 套)") # ─── 追问清单 ────────────────────────────────────────────────────────── async def _seed_inquiries(db: AsyncSession) -> None: """月报追问清单 — AI 生成补充问题。""" inquiries = [ InquiryList(company_id=COMP_A, questions=[{"q": "本月流失客户的具体原因是什么？", "a": "客户反馈技术支持响应慢，竞品以低价策略抢客。"}, {"q": "AI推理成本降低12%的具体措施？", "a": "采用模型量化+推理优化，单次调用成本从0.05元降至0.044元。"}, {"q": "新签3家客户的年合同金额？", "a": "合计约480万/年，平均160万/家。"}], status="answered", sent_at=D(9), answered_at=D(5)), InquiryList(company_id=COMP_B, questions=[{"q": "营收连续2个月持平的原因？", "a": "现有客户续约稳定，但新客户拓展速度放缓。"}, {"q": "金融客户拓展进展？", "a": "1家银行已进POC阶段，预计Q3签约。"}, {"q": "A+轮融资准备情况？", "a": "BP初稿完成，正在整理财务预测。"}], status="answered", sent_at=D(7), answered_at=D(3)), InquiryList(company_id=COMP_D, questions=[{"q": "NPU原型流片时间表？", "a": "预计下月流片，Q4出工程样片。"}, {"q": "天使+轮BP何时可以准备好？", "a": "正在准备中，预计2周内完成。"}], status="sent", sent_at=D(2)), ] for i in inquiries: db.add(i) await db.flush() print(" ✓ 追问清单 (3 条)") # ─── Agent 执行记录 ──────────────────────────────────────────────────── async def _seed_agent_executions(db: AsyncSession) -> None: """AI Agent L1-L4 分级执行记录。""" agents = [ AgentExecution(tenant_id=TENANT_ID, agent_name="月报解析Agent", autonomy_level="L3", input_summary="智链科技2025-06月报", output_summary="解析完成：营收800万(+15%)，客户净增2家", output_detail={"revenue": 800, "mom_growth": 0.15, "new_customers": 3, "churned": 1}, review_status="auto_approved", model_version="gpt-4o-2024", duration_ms=3200), AgentExecution(tenant_id=TENANT_ID, agent_name="风险预警Agent", autonomy_level="L2", input_summary="量子芯微财务数据扫描", output_summary="发现1项critical风险：Runway不足7月", output_detail={"risks_found": 1, "severity": "critical", "metric": "runway", "value": 7}, review_status="approved", reviewer_id=MGR_ID, reviewed_at=D(3), model_version="gpt-4o-2024", duration_ms=1800), AgentExecution(tenant_id=TENANT_ID, agent_name="弱信号监测Agent", autonomy_level="L3", input_summary="深瞳智能外部信号扫描", output_summary="LinkedIn检测到CTO更新简历", output_detail={"signal_type": "org", "source": "LinkedIn", "confidence": 0.85}, review_status="auto_approved", model_version="gpt-4o-2024", duration_ms=2400), AgentExecution(tenant_id=TENANT_ID, agent_name="协同匹配Agent", autonomy_level="L2", input_summary="Portfolio协同分析", output_summary="发现3个协同机会", output_detail={"matches": 3, "types": ["customer", "tech", "talent"]}, review_status="approved", reviewer_id=LEAD_ID, reviewed_at=D(5), model_version="gpt-4o-2024", duration_ms=5600), AgentExecution(tenant_id=TENANT_ID, agent_name="健康度计算Agent", autonomy_level="L4", input_summary="5家企业健康度季度评估", output_summary="14维度评分完成，2家上升1家下降", output_detail={"companies": 5, "trends": {"up": 2, "stable": 2, "down": 1}}, review_status="auto_approved", model_version="gpt-4o-2024", duration_ms=8200), AgentExecution(tenant_id=TENANT_ID, agent_name="月报解析Agent", autonomy_level="L3", input_summary="云栈数据2025-06月报", output_summary="解析完成：营收150万(持平)", output_detail={"revenue": 150, "mom_growth": 0.0}, review_status="auto_approved", model_version="gpt-4o-2024", duration_ms=2800), AgentExecution(tenant_id=TENANT_ID, agent_name="决策哨兵Agent", autonomy_level="L2", input_summary="云栈数据融资时机分析", output_summary="建议现在启动A+轮融资", output_detail={"decision": "funding", "recommendation": "now", "confidence": 0.72}, review_status="pending", model_version="gpt-4o-2024", duration_ms=4200), AgentExecution(tenant_id=TENANT_ID, agent_name="数据校验Agent", autonomy_level="L4", input_summary="5家企业财务数据交叉校验", output_summary="校验通过，数据一致性98.5%", output_detail={"checked": 30, "passed": 29, "inconsistencies": 1}, review_status="auto_approved", model_version="gpt-4o-2024", duration_ms=1500), ] for idx, a in enumerate(agents): a.created_at = D(10 - idx) db.add(a) await db.flush() print(" ✓ Agent 执行记录 (8 条)") # ─── 客户获取计划 ────────────────────────────────────────────────────── async def _seed_customer_plans(db: AsyncSession) -> None: """AI 客户增长引擎 — LP 资源匹配 + 客户获取方案。""" plans = [ CustomerAcquisitionPlan(company_id=COMP_A, target_customer="年营收10-50亿制造业CIO/供应链总监", entry_angle="从供应链预测场景切入，ROI 3个月可见", decision_chain={"roles": ["CIO", "供应链VP", "采购总监"], "influence": {"CIO": 0.4, "供应链VP": 0.4, "采购总监": 0.2}, "cycle": "3-6月"}, pricing_strategy="SaaS订阅：年费80-200万，按模块阶梯定价", competitive_analysis={"direct": ["杉数科技", "京东物流AI"], "advantage": "多模态决策引擎，部署快3倍", "weakness": "品牌知名度低"}, lp_resources={"lp_company": "某制造业LP", "intro_channel": "GP引荐", "warmth": "high"}, execution_status="executing", result={"customers_contacted": 5, "poc_started": 2, "signed": 1, "revenue": 160}), CustomerAcquisitionPlan(company_id=COMP_B, target_customer="金融机构数据治理负责人", entry_angle="合规驱动+AI训练数据管理，从数据治理切入", decision_chain={"roles": ["数据治理总监", "合规负责人", "CTO"], "influence": {"数据治理总监": 0.5, "合规负责人": 0.3, "CTO": 0.2}, "cycle": "6-12月"}, pricing_strategy="平台授权+实施服务：首年120万+实施80万", competitive_analysis={"direct": ["亿信华辰", "数美科技"], "advantage": "AI原生，支持大模型训练数据管理", "weakness": "金融行业案例少"}, lp_resources={"lp_company": "某银行LP", "intro_channel": "GP引荐", "warmth": "medium"}, execution_status="planned", result=None), CustomerAcquisitionPlan(company_id=COMP_C, target_customer="半导体晶圆厂质检负责人", entry_angle="替代人工目检，从良率提升切入", decision_chain={"roles": ["质量总监", "生产VP", "CIO"], "influence": {"质量总监": 0.5, "生产VP": 0.3, "CIO": 0.2}, "cycle": "3-6月"}, pricing_strategy="设备+软件：单条产线200-400万", competitive_analysis={"direct": ["康代智能", "奥普特"], "advantage": "AI算法精度99.2%，行业领先", "weakness": "硬件成本偏高"}, lp_resources={"lp_company": "某半导体LP", "intro_channel": "LP直接引荐", "warmth": "high"}, execution_status="completed", result={"customers_contacted": 8, "poc_started": 5, "signed": 3, "revenue": 800}), ] for p in plans: db.add(p) await db.flush() print(" ✓ 客户获取计划 (3 条)") # ─── 退出预测 ────────────────────────────────────────────────────────── async def _seed_exit_predictions(db: AsyncSession) -> None: """退出时机预测 — IPO/并购/二手份额。""" predictions = [ ExitPrediction(company_id=COMP_C, exit_path="ipo", timing_window={"earliest": "2027-Q1", "latest": "2028-Q3"}, expected_return=4.2, hold_return=2.8, confidence=0.75, signals={"positive": ["营收年增50%+", "半导体赛道热度高", "专利壁垒"], "negative": ["客户集中度60%", "CTO稳定性"]}, recommendation="建议2027年Q1启动IPO准备，当前应解决客户集中度和CTO retention问题。"), ExitPrediction(company_id=COMP_A, exit_path="acquisition", timing_window={"earliest": "2026-Q3", "latest": "2027-Q4"}, expected_return=3.5, hold_return=3.0, confidence=0.65, signals={"positive": ["多模态技术差异化", "制造业客户基础"], "negative": ["客户流失率上升", "竞争加剧"]}, recommendation="建议关注产业并购机会，优先接触京东物流、菜鸟等供应链平台。"), ExitPrediction(company_id=COMP_E, exit_path="acquisition", timing_window={"earliest": "2028-Q1", "latest": "2030-Q4"}, expected_return=5.0, hold_return=3.5, confidence=0.55, signals={"positive": ["AI药物发现赛道升温", "3靶点进入先导"], "negative": ["研发周期长", "现金流压力"]}, recommendation="建议继续持有，等待靶点临床验证结果后评估并购机会。"), ] for p in predictions: db.add(p) await db.flush() print(" ✓ 退出预测 (3 条)") # ─── Pre-mortem + Red Team ───────────────────────────────────────────── async def _seed_pre_mortems(db: AsyncSession) -> None: """失败推演 + 对抗分析。""" pre_mortems = [ PreMortemRecord(company_id=COMP_D, decision_context="天使+轮融资1500万，估值8000万", failure_paths=[{"path": "流片失败导致估值缩水", "probability": 0.25, "impact": "high"}, {"path": "融资周期超3月，现金耗尽", "probability": 0.15, "impact": "critical"}, {"path": "竞品发布同类NPU，差异化丧失", "probability": 0.20, "impact": "medium"}], risk_checklist={"items": ["流片进度", "现金Runway", "竞品动态", "核心团队稳定性"], "checked": 3, "total": 4}, mitigations={"actions": ["准备过桥贷款方案", "与2家竞品做技术对标", "CTO股权激励绑定"]}), PreMortemRecord(company_id=COMP_B, decision_context="A+轮融资3000万 vs 等待Q2数据验证", failure_paths=[{"path": "现在融资估值偏低，稀释过多", "probability": 0.40, "impact": "medium"}, {"path": "等待期间现金耗尽", "probability": 0.20, "impact": "critical"}, {"path": "Q2数据不及预期，融资更难", "probability": 0.30, "impact": "high"}], risk_checklist={"items": ["现金Runway", "Q2营收预期", "竞品定价压力", "金融客户签约"], "checked": 2, "total": 4}, mitigations={"actions": ["准备两套BP", "与现有股东沟通过桥", "加速金融客户POC"]}), ] for p in pre_mortems: db.add(p) red_teams = [ RedTeamRecord(company_id=COMP_A, perspective="competitor", analysis="如果我是杉数科技，会以低价+行业深耕策略抢智链科技的制造业客户。智链的多模态优势在具体场景中并未形成壁垒。", vulnerabilities={"areas": ["客户成功团队薄弱", "品牌知名度低", "定价偏高"], "severity": "medium"}, counterarguments={"defense": ["多模态技术差异化", "客户切换成本高"]}), RedTeamRecord(company_id=COMP_C, perspective="pessimistic_investor", analysis="深瞳智能半导体占比60%是定时炸弹。半导体周期下行时营收将大幅缩水。CTO稳定性问题如果恶化，技术优势可能丧失。", vulnerabilities={"areas": ["客户集中度60%", "CTO稳定性0.65", "半导体周期风险"], "severity": "high"}, counterarguments={"defense": ["正在拓展新能源客户", "CTO激励方案设计中"]}), RedTeamRecord(company_id=COMP_D, perspective="devils_advocate", analysis="量子芯微的NPU在边缘端确实有需求，但巨头(英伟达/高通)一旦下沉，创业公司很难竞争。7月Runway意味着没有试错空间。", vulnerabilities={"areas": ["巨头竞争风险", "Runway 7月", "无营收"], "severity": "critical"}, counterarguments={"defense": ["边缘端低功耗是差异化", "已有3家潜在客户"]}), ] for r in red_teams: db.add(r) await db.flush() print(" ✓ Pre-mortem (2) + Red Team (3)") # ─── 产品竞争力诊断 ──────────────────────────────────────────────────── async def _seed_product_diagnostics(db: AsyncSession) -> None: """产品竞争力诊断 — 热力图 + 竞品对比。""" diagnostics = [ ProductDiagnostic(company_id=COMP_A, product_name="多模态供应链决策引擎", dimensions={"技术壁垒": 82, "产品成熟度": 75, "用户体验": 70, "生态集成": 65, "定价竞争力": 60, "客户支持": 55}, heatmap_data={"strengths": ["技术壁垒", "产品成熟度"], "weaknesses": ["客户支持", "定价竞争力"], "neutral": ["用户体验", "生态集成"]}, competitors=[{"name": "杉数科技", "scores": {"技术壁垒": 78, "产品成熟度": 85, "用户体验": 80, "生态集成": 75, "定价竞争力": 70, "客户支持": 82}}, {"name": "京东物流AI", "scores": {"技术壁垒": 70, "产品成熟度": 88, "用户体验": 85, "生态集成": 90, "定价竞争力": 65, "客户支持": 78}}], roadmap_suggestions="建议优先提升客户支持能力（增加2名CSM）和优化定价策略（引入阶梯定价），同时加强生态集成（对接ERP系统）。"), ProductDiagnostic(company_id=COMP_C, product_name="工业质检视觉AI平台", dimensions={"技术壁垒": 88, "产品成熟度": 85, "用户体验": 78, "生态集成": 72, "定价竞争力": 65, "客户支持": 80}, heatmap_data={"strengths": ["技术壁垒", "产品成熟度", "客户支持"], "weaknesses": ["定价竞争力", "生态集成"], "neutral": ["用户体验"]}, competitors=[{"name": "康代智能", "scores": {"技术壁垒": 82, "产品成熟度": 90, "用户体验": 82, "生态集成": 85, "定价竞争力": 75, "客户支持": 85}}, {"name": "奥普特", "scores": {"技术壁垒": 75, "产品成熟度": 82, "用户体验": 78, "生态集成": 80, "定价竞争力": 82, "客户支持": 80}}], roadmap_suggestions="技术优势明显，建议降低硬件成本（考虑国产替代方案）并加强生态集成（对接MES系统）。"), ] for d in diagnostics: db.add(d) await db.flush() print(" ✓ 产品诊断 (2 条)") # ─── 同行学习圈 ──────────────────────────────────────────────────────── async def _seed_peer_circles(db: AsyncSession) -> None: """同行学习圈 — AI 匹配面临类似挑战的创始人。""" circles = [ PeerLearningCircle(tenant_id=TENANT_ID, topic="融资时机选择：现在 vs 等待验证", description="云栈数据和量子芯微创始人均面临融资时机决策，AI匹配讨论。", members=[{"founder_id": FOUNDER_B, "company_id": COMP_B, "name": "刘云栈"}, {"founder_id": FOUNDER_C, "company_id": COMP_D, "name": "赵深瞳(代)"}], discussion_framework={"steps": ["各自分享融资困境", "分析共同风险", "讨论应对策略", "制定行动承诺"]}, conclusions="两位创始人一致认为：在营收未达预期时，应优先保证现金Runway > 12月，融资宁可早不要晚。", action_commitments=[{"founder": "刘云栈", "action": "2周内完成BP", "deadline": "14天"}, {"founder": "赵深瞳(代)", "action": "启动天使+轮", "deadline": "7天"}], status="completed"), PeerLearningCircle(tenant_id=TENANT_ID, topic="客户集中度风险管理", description="深瞳智能和智链科技都面临客户集中度/流失问题。", members=[{"founder_id": FOUNDER_A, "company_id": COMP_A, "name": "陈智链"}, {"founder_id": FOUNDER_C, "company_id": COMP_C, "name": "赵深瞳"}], discussion_framework={"steps": ["分享客户结构", "分析流失原因", "讨论多元化策略", "制定行动承诺"]}, conclusions="深瞳：半导体占比60%需降低；智链：流失率4%需降至3%以下。共同策略：建立客户健康度预警机制。", action_commitments=[{"founder": "陈智链", "action": "建立客户健康度周报", "deadline": "30天"}, {"founder": "赵深瞳", "action": "启动新能源客户拓展", "deadline": "60天"}], status="active"), ] for c in circles: db.add(c) await db.flush() print(" ✓ 同行学习圈 (2 个)") # ─── 数字孪生 ────────────────────────────────────────────────────────── async def _seed_digital_twins(db: AsyncSession) -> None: """数字孪生模型。""" twins = [ DigitalTwinModel(company_id=COMP_A, model_params={"revenue_model": "saaS_subscription", "churn_rate": 0.03, "growth_rate": 0.15, "cac": 50, "ltv": 400}, scenarios=[{"name": "乐观", "revenue_y0": 800, "revenue_y1": 1200, "runway_months": 18}, {"name": "基准", "revenue_y0": 800, "revenue_y1": 1000, "runway_months": 14}, {"name": "悲观", "revenue_y0": 800, "revenue_y1": 700, "runway_months": 10}], accuracy_score=0.82, last_calibrated_at=D(7)), DigitalTwinModel(company_id=COMP_C, model_params={"revenue_model": "hardware+software", "semi_ratio": 0.60, "growth_rate": 0.22, "gross_margin": 0.55}, scenarios=[{"name": "乐观", "revenue_y0": 1200, "revenue_y1": 1800, "valuation": "15亿"}, {"name": "基准", "revenue_y0": 1200, "revenue_y1": 1500, "valuation": "12亿"}, {"name": "悲观", "revenue_y0": 1200, "revenue_y1": 1000, "valuation": "8亿"}], accuracy_score=0.78, last_calibrated_at=D(14)), ] for t in twins: db.add(t) await db.flush() print(" ✓ 数字孪生 (2 个)") # ─── 干预事件 + 结果（Alpha 归因） ───────────────────────────────────── async def _seed_interventions(db: AsyncSession) -> None: """投后管理 Alpha 归因 — 干预 → 指标变化 → 估值影响。""" interventions = [ InterventionEvent(company_id=COMP_A, intervention_type="customer_intro", title="引荐LP制造业客户给智链科技", description="通过LP关系引荐3家制造业客户，加速客户拓展。", executed_by=GP_ID, executed_at=D(30)), InterventionEvent(company_id=COMP_C, intervention_type="strategy", title="建议深瞳智能拓展新能源客户", description="降低半导体客户集中度，拓展新能源赛道。", executed_by=LEAD_ID, executed_at=D(45)), InterventionEvent(company_id=COMP_A, intervention_type="governance", title="推动智链科技建立客户成功团队", description="增加2名CSM，建立客户健康度周报机制。", executed_by=MGR_ID, executed_at=D(60)), InterventionEvent(company_id=COMP_D, intervention_type="funding", title="协助量子芯微准备天使+轮融资", description="GP直接参与BP打磨和投资人对接。", executed_by=GP_ID, executed_at=D(10)), ] for i in interventions: db.add(i) await db.flush() results = [ InterventionResult(intervention_id=interventions[0].id, metric_changes={"customers_new": 2, "revenue_increase": 320, "sales_cycle_reduction": 30}, valuation_impact=0.5, return_contribution=0.15, alpha_score=0.82, evidence={"before": {"customers": 48, "revenue": 800}, "after": {"customers": 50, "revenue": 960}}), InterventionResult(intervention_id=interventions[1].id, metric_changes={"semi_ratio": -0.05, "new_customers": 1, "revenue_diversification": 0.10}, valuation_impact=0.3, return_contribution=0.08, alpha_score=0.68, evidence={"before": {"semi_ratio": 0.65, "customers": 20}, "after": {"semi_ratio": 0.60, "customers": 21}}), InterventionResult(intervention_id=interventions[2].id, metric_changes={"churn_rate": -0.01, "csat": 8, "response_time": -40}, valuation_impact=0.2, return_contribution=0.06, alpha_score=0.72, evidence={"before": {"churn_rate": 0.04, "csat": 7.2}, "after": {"churn_rate": 0.03, "csat": 8.0}}), ] for r in results: db.add(r) await db.flush() print(" ✓ 干预事件 (4) + 结果 (3) — Alpha 归因") # ─── 组合再平衡 + Monte Carlo ────────────────────────────────────────── async def _seed_portfolio_simulations(db: AsyncSession) -> None: """组合再平衡建议 + Monte Carlo 模拟结果。""" db.add(PortfolioRebalancing( tenant_id=TENANT_ID, marginal_returns={"COMP_A": 0.25, "COMP_B": 0.08, "COMP_C": 0.35, "COMP_D": -0.05, "COMP_E": 0.15}, reallocation_plan={"reduce": [{"company": "COMP_B", "amount": 500, "reason": "边际回报低"}], "increase": [{"company": "COMP_C", "amount": 300, "reason": "高增长高回报"}, {"company": "COMP_A", "amount": 200, "reason": "上升趋势明确"}]}, irr_impact=0.03, dpi_impact=0.02, status="proposed", )) # Monte Carlo — 生成 IRR/DPI 概率分布 random.seed(42) irr_samples = [random.gauss(0.22, 0.08) for _ in range(10000)] irr_buckets = {} for s in irr_samples: bucket = round(s * 2) / 2 # 0.5% 区间 irr_buckets[str(bucket)] = irr_buckets.get(str(bucket), 0) + 1 dpi_samples = [max(0, random.gauss(1.8, 0.4)) for _ in range(10000)] dpi_buckets = {} for s in dpi_samples: bucket = round(s * 2) / 2 dpi_buckets[str(bucket)] = dpi_buckets.get(str(bucket), 0) + 1 db.add(MonteCarloSimulation( tenant_id=TENANT_ID, iterations=10000, irr_distribution=irr_buckets, dpi_distribution=dpi_buckets, percentile_p5=0.08, percentile_p50=0.22, percentile_p95=0.36, )) await db.flush() print(" ✓ 组合再平衡 (1) + Monte Carlo (1)") # ─── 知识图谱 ────────────────────────────────────────────────────────── async def _seed_knowledge_graph(db: AsyncSession) -> None: """知识图谱节点 — 企业特征 + 管理动作 + 环境上下文 → 结果 → 回报影响。""" nodes = [ KnowledgeNode(tenant_id=TENANT_ID, entity_type="company", entity_id=COMP_A, attributes={"name": "智链科技", "industry": "AI/供应链", "stage": "B"}, relations=[{"target": "action_1", "type": "received"}, {"target": "context_1", "type": "in"}]), KnowledgeNode(tenant_id=TENANT_ID, entity_type="action", entity_id="action_1", attributes={"type": "customer_intro", "title": "引荐LP制造业客户"}, relations=[{"target": "result_1", "type": "led_to"}]), KnowledgeNode(tenant_id=TENANT_ID, entity_type="context", entity_id="context_1", attributes={"type": "market", "description": "制造业数字化转型加速"}, relations=[{"target": "result_1", "type": "influenced"}]), KnowledgeNode(tenant_id=TENANT_ID, entity_type="result", entity_id="result_1", attributes={"metric": "revenue", "change": "+20%", "duration": "3月"}, relations=[{"target": "return_1", "type": "contributed_to"}]), KnowledgeNode(tenant_id=TENANT_ID, entity_type="return", entity_id="return_1", attributes={"irr_impact": 0.15, "alpha_score": 0.82}, relations=[]), ] for n in nodes: db.add(n) await db.flush() print(" ✓ 知识图谱 (5 节点)") # ─── 数据源 ──────────────────────────────────────────────────────────── async def _seed_data_sources(db: AsyncSession) -> None: """外部数据源配置。""" sources = [ DataSource(tenant_id=TENANT_ID, source_type="crunchbase", name="Crunchbase API", api_endpoint="https://api.crunchbase.com/v3.1", status="active", last_synced_at=D(1), config={"sync_frequency": "daily"}), DataSource(tenant_id=TENANT_ID, source_type="github", name="GitHub 活跃度监测", api_endpoint="https://api.github.com", status="active", last_synced_at=D(1), config={"repos": ["zhilian/ai-engine", "yunzhan/data-platform"]}), DataSource(tenant_id=TENANT_ID, source_type="business_registry", name="工商信息同步", api_endpoint="https://api.qcc.com/v2", status="active", last_synced_at=D(3), config={"sync_frequency": "weekly"}), DataSource(tenant_id=TENANT_ID, company_id=COMP_C, source_type="custom", name="半导体行业数据库", api_endpoint="https://semi-data.example.com/api", status="active", last_synced_at=D(2), config={"metrics": ["industry_growth", "competitor_funding"]}), ] for s in sources: db.add(s) await db.flush() print(" ✓ 数据源 (4 个)") if __name__ == "__main__": asyncio.run(seed())