Welcome to the Venture Capital Intelligence Toolkit — a purpose-built repository that transforms raw data into actionable investment insights. Think of this as a co-pilot for venture analysts, angel investors, and fund managers who need to move faster without sacrificing depth. This isn't another dashboard wrapper; it's a modular intelligence system that combines Claude's reasoning capabilities with Python's deterministic power to deliver nine distinct venture skills, from screening pre-seed startups to modeling complex cap tables.
- Inspiration & Philosophy
- Architecture Overview (Mermaid Diagram)
- Core Skills Breakdown
- Installation & Configuration
- Example: Profile Configuration
- Example: Console Invocation
- Key Features
- API Integration: Claude & OpenAI
- Multilingual Support & Responsive UI
- Emoji OS Compatibility Table
- SEO-Optimized Keyword Integration
- Disclaimer
- License
The original context of this repository—venture capital intelligence—evokes a world where decisions are only as good as the data feeding them. Yet most analysts rely on disjointed tools: one for market sizing, another for financial modeling, a third for cap table math. This fragmentation breeds inconsistency. The Venture Capital Intelligence Toolkit merges these into a single pipeline where Claude reason and Python precision work in concert. The result? A system that feels less like software and more like a sixth sense for evaluating startups, markets, and financial structures.
graph TD
A[Raw Data Input] --> B[Claude Reasoning Engine]
A --> C[Python Deterministic Modules]
B --> D[Startup Screening]
B --> E[Qualitative Assessment]
B --> F[Trend Identification]
C --> G[Financial Modeling]
C --> H[Market Sizing]
C --> I[Cap Table Computation]
D --> J[Unified Intelligence Output]
E --> J
F --> J
G --> J
H --> J
I --> J
J --> K[Responsive UI Layer]
J --> L[API Export]
J --> M[Console Reports]
K --> N[Multilingual Dashboards]
L --> O[Third-Party Integration]
This architecture reflects a dual-engine design: Claude handles the ambiguous, human-like reasoning tasks (screening, trends, qualitative depth), while Python executes deterministic calculations (modeling, sizing, cap table math). They converge into a single intelligence output that can be consumed via a responsive UI, API, or console.
Nine skills form the backbone of this toolkit. Three run exclusively on Claude's reasoning; six combine Claude with Python for deterministic outputs.
- Startup Screening — Claude analyzes pitch decks, founder backgrounds, and market fit using pattern recognition trained on thousands of successful and failed ventures.
- Qualitative Assessment — Evaluates team dynamics, competitive moats, and execution capability using nuanced language understanding.
- Trend Identification — Scans news, patents, and funding announcements to surface emerging sectors before they become crowded.
- Financial Modeling — Python calculates revenue projections, burn rates, and unit economics; Claude contextualizes assumptions and flags anomalies.
- Market Sizing — Python aggregates data from multiple sources; Claude validates methodology and provides narrative context.
- Cap Table Computation — Python performs precise equity dilution and waterfall analysis; Claude explains implications for founders and investors.
- Risk Scoring — Python normalizes quantitative risks; Claude adjusts weights based on qualitative inputs.
- Comparative Valuation — Python comps multiples and benchmarks; Claude interprets relative positioning.
- Portfolio Optimization — Python runs Monte Carlo simulations; Claude recommends allocation strategies based on risk appetite.
- Python 3.10+
- Claude API key (Anthropic)
- OpenAI API key (optional, for alternative language models)
- Git installed on your system
-
Clone the repository:
git clone https://github.com/your-username/venture-capital-intelligence.git cd venture-capital-intelligence -
Install dependencies:
pip install -r requirements.txt
-
Configure environment variables:
export CLAUDE_API_KEY="your-claude-api-key" export OPENAI_API_KEY="your-openai-api-key" # optional
-
Run the first-time setup:
python setup.py --initialize
Configuration profiles let you define your investment thesis, risk tolerance, and preferred sectors. Below is a sample profile for a seed-stage fund focusing on climate tech:
# profile_climate_seed.yaml
fund_name: "Green Horizon Ventures"
stage: "pre-seed to seed"
sectors: ["climate tech", "carbon removal", "sustainable agriculture"]
geography: "North America, Europe"
check_size_min: 250000
check_size_max: 2000000
risk_tolerance: "moderate-high"
required_metrics:
- t2d3_rule: true
- gross_margin_target: 0.6
- founder_domain_expertise: "required"
exclude_sectors: ["crypto", "gambling", "tobacco"]
llm_preference: "claude-3-opus"
output_format: "markdown_report"This profile feeds directly into the screening engine. When a new startup enters the pipeline, Claude references these parameters to assess alignment.
Here's how you might run the toolkit from the command line to screen a startup and generate a comprehensive report:
python scout.py \
--profile profile_climate_seed.yaml \
--pitchdeck "startups/carbon_capture_v1.pdf" \
--financials "data/revenue_carbon_capture.csv" \
--market "carbon_removal_market_2026.json" \
--output "report_climate_v1.md" \
--format markdown \
--verboseThis command triggers the full pipeline: Claude screens the pitch deck, Python models the financials, market sizing runs on imported data, and a final report is generated in Markdown. Adding --verbose shows each skill's contribution in real-time.
- Dual-Engine Intelligence: Claude handles reasoning tasks (screening, trends, qualitative analysis) while Python manages deterministic computations (cap tables, financial models, market sizing).
- Responsive UI: Built with a lightweight web interface that adapts to desktop, tablet, and mobile views. Access reports from anywhere.
- Multilingual Support: Reports and dashboards available in English, Mandarin, Spanish, French, German, and Japanese. Claude's language capability enables translation without degrading analytical quality.
- 24/7 Automated Screening: The pipeline runs continuously, ingesting new startup submissions and generating preliminary assessments within minutes.
- Customizable Risk Profiles: Define your investment thesis across sectors, stages, geographies, and risk levels. The toolkit respects these boundaries.
- Export Flexibility: Results can be exported as Markdown, PDF, JSON, or CSV. Integrate with your existing tools via the API layer.
- Real-Time Data Feeds: Connect to Crunchbase, PitchBook, or custom APIs to keep market sizing and trend identification current.
- Audit Trails: Every decision includes a reasoning trace from Claude and raw calculations from Python, ensuring transparency.
- Scalable Architecture: Run locally for individual analysis or deploy on cloud infrastructure for team-wide use.
This toolkit supports both Claude (Anthropic) and OpenAI (GPT-4 / GPT-4o) as the language model backend. Claude is the default due to its superior reasoning depth for venture-related analysis, but OpenAI can be used as a fallback.
- Default model: Claude 3 Opus (recommended for complex reasoning)
- Rate limits: Adjustable via environment variables
- Use case: Qualitative screening, trend analysis, narrative generation
- Default model: GPT-4o
- Use case: Alternative reasoning engine, multilingual translation, and when Claude is unavailable
Configuration:
from vc_intelligence import VCIntelligence
vc = VCIntelligence(llm="claude-3-opus", api_key="your-key")
result = vc.screen_startup("path/to/pitchdeck.pdf")| Language | Status | Quality Rating |
|---|---|---|
| English | Full | Native |
| Mandarin | Full | Near-native |
| Spanish | Full | Near-native |
| French | Full | Native |
| German | Full | Near-native |
| Japanese | Beta | Good |
The toolkit's web interface uses a flexible grid system that collapses to a single column on mobile devices. Charts and tables resize automatically. The console output remains unchanged across platforms.
Automated screening runs around the clock with no human intervention. Output reports are generated and stored locally or pushed to cloud storage. The system includes error handling for API timeouts, missing data, and malformed inputs.
| Emoji | Windows | macOS | Linux | Android | iOS |
|---|---|---|---|---|---|
| 🚀 | Full | Full | Partial | Full | Full |
| 💰 | Full | Full | Full | Full | Full |
| 📊 | Full | Full | Full | Full | Full |
| 🔍 | Full | Full | Partial | Full | Full |
| 🛡️ | Full | Full | Partial | Full | Full |
| ⚡ | Full | Full | Full | Full | Full |
| 🌐 | Full | Full | Full | Full | Full |
| 🧠 | Full | Full | Partial | Full | Full |
| 📈 | Full | Full | Full | Full | Full |
| 🔄 | Full | Full | Full | Full | Full |
Note: Linux emoji rendering depends on installed fonts. Install noto-fonts-emoji for full support.
This repository targets keywords that venture capital professionals and startup analysts search for:
- Venture capital intelligence tools
- Startup screening automation
- Cap table calculator software
- Market sizing for VCs
- AI-powered investment analysis
- Financial modeling for early-stage startups
- Claude API for venture capital
- Python venture capital toolkit
- Automated investment memo generation
These terms appear naturally in the documentation, code comments, and example outputs. The system is designed to solve real problems—not just rank for keywords.
Important: The Venture Capital Intelligence Toolkit is an analytical aid, not a replacement for professional investment judgment. All outputs—including screening scores, financial projections, market size estimates, and risk assessments—should be reviewed by qualified professionals before making investment decisions. Past performance of the reasoning engine does not guarantee future results. The creators assume no liability for financial losses or missed opportunities arising from the use of this software. Always consult with legal, financial, and tax advisors when evaluating venture investments.
This project is licensed under the MIT License. See the LICENSE file for full terms.
Built for analysts who want to see around corners. Version 2.0 — 2026.