Multi-agent finance simulation for educational portfolios
Educational simulation only — not financial advice.
No real trading, no brokerage connection, no live market data. Every case below is a
preset toy scenario built to demonstrate multi-agent reasoning, not a recommendation to buy or sell anything.
⚙️ This is a deterministic, rule-based simulator inspired by multi-agent investment-committee
workflows — not a real LLM agent system. Every score comes from a fixed formula, so the whole
decision is reproducible and auditable (see “How this decision was calculated” below).
1 · Choose a case
Select a case to enable the committee.
2 · Committee deliberation
3 · Final committee decision
4 · How this decision was calculated
What you can put on your resume
Built an AI Investment Committee Simulator: a deterministic multi-agent simulator where four analyst/risk roles score a scenario and a portfolio-manager coordinator aggregates their weighted signals into one explainable decision.
Designed a transparent decision engine (weighted signal aggregation + confidence from inter-agent agreement) so every verdict is fully traceable to the underlying agent stances — no black box.
Shipped a zero-dependency, offline-first cockpit (static HTML + a mirrored Python CLI) with strong educational/safety framing and preset scenarios instead of live trading.
See CAREER_KIT.md for the full STAR interview story and five follow-up Q&A.