Silicon-Valley

Market Council — Intelligence With Boundaries Category: Agents You Can Audit Tech Stack: MongoDB Atlas, Node.js, React, Finnhub API, Docker Project Description What if AI agents could make investment decisions for us, but couldn't act without accountability? Market Council is a multi-agent investment intelligence platform that combines collaborative AI research with enterprise-grade governance, policy enforcement, and auditability powered by MongoDB Atlas. Instead of relying on a single AI model, Market Council coordinates specialized multiple agents responsible for market monitoring, financial news analysis, quantitative research, risk assessment, and portfolio management. These agents collaborate to analyze multiple stocks and ETFs, producing explainable **BUY, HOLD, and SELL signals** using market data from Finnhub. But intelligence alone isn't enough, especially in financial services. Every proposed investment action must pass through a MongoDB Atlas-backed governance layer before execution. Our platform implements three core principles: - Identity: Every agent has a defined identity and editable permissions stored in MongoDB Atlas. Unauthorized agents cannot perform restricted actions. - Policy: Investment actions are checked against configurable risk limits, portfolio concentration rules, and approval requirements before execution. Policies are versioned and enforced through deterministic authorization checks. - Auditability: Agent communications, recommendations, authorization decisions, human approvals, and simulated brokerage outcomes are recorded in MongoDB Atlas, allowing reviewers to reconstruct what happened and why. MongoDB Change Streams support real-time revocation through an emergency kill switch, enabling operators to stop pending actions when permissions or policies change. Market Council also integrates a separate simulated brokerage service to demonstrate real API execution, order verification, and cancellation without risking actual funds. What Makes Market Council Different? Most AI investment tools focus on generating recommendations. Market Council focuses on making autonomous financial agents accountable. Our agents can collaborate, disagree, and recommend actions, but they cannot bypass defined permissions, financial controls, or human oversight. By combining investment intelligence with MongoDB-powered identity, policy enforcement, and transparent auditing, Market Council demonstrates how agentic AI can move beyond experimental chatbots toward controlled, trustworthy automation in financial services. Our vision: AI agents that are intelligent enough to recommend, governed enough to trust, and transparent enough to audit. Tagline: Agents recommend. Policies decide. MongoDB proves.

Team