An AI-assisted strategy tool for answering market entry questions with sourced research, decision thresholds, and a verification pass.
This project is designed to simulate a consulting-style market entry workstream: validate the question, research the market, produce a structured recommendation, check factual claims against sources, and return a transparent executive report.
- Validates that the user provided a specific company and market.
- Researches market size, competitors, risks, and entry barriers.
- Produces a six-section market entry report.
- Requires a clear verdict:
Enter,Don't Enter, orEnter with conditions. - Adds quantified decision thresholds so the recommendation is testable.
- Runs a verification pass to flag unsupported claims and repair weak sections.
- Appends a transparency footer showing source count, checked claims, and corrections.
See docs/sample_report.md for the expected report structure and decision logic.
python scripts/validate.pyThe core parser tests run offline. The adversarial tests are skipped unless ANTHROPIC_API_KEY is configured.
-
User enters a market entry question, such as:
Should Spotify enter the podcast advertising market in Southeast Asia? -
The app validates that the question includes:
- a specific company
- a specific market, geography, industry, or segment
-
The analysis pipeline creates a sourced report covering:
- executive summary
- market overview
- competitive landscape
- risks and entry barriers
- decision thresholds
- recommendation detail
-
A verification step checks numeric claims, competitor facts, verdict consistency, and decision thresholds.
Market entry recommendations often fail because they mix assumptions, unsourced claims, and vague conclusions. This tool forces a more disciplined structure:
- every major number needs source support
- missing data must be stated instead of invented
- the final recommendation must match the executive verdict
- decision thresholds make the recommendation measurable
- Python
- Streamlit
- Anthropic Claude API
- Web search tool use
- Pytest
python3 -m venv .venv
source .venv/bin/activate
python -m pip install -r requirements.txt
cp .env.example .envAdd your API key to .env:
ANTHROPIC_API_KEY=your-api-key-hereThen run:
streamlit run app.pyThe test suite includes adversarial checks for vague prompts, prompt injection, unsupported claims, contradiction handling, and low-data markets.
Some tests require a live ANTHROPIC_API_KEY because they exercise the full research and verification pipeline.
pytest tests/test_adversarial.py -vWithout an API key, the live tests are skipped.
Should Netflix enter the live sports streaming market in Brazil?
Should IKEA enter the modular housing market in Japan?
Should Starbucks enter the energy drink market in Germany?
Built an AI-powered Market Entry Strategy Analyzer that validates business questions, performs sourced market research, checks claims against evidence, and generates consulting-style recommendations with decision thresholds.
This tool relies on current web research and model-generated analysis. Outputs should be reviewed before being used for real investment, expansion, or commercial decisions.