Fintech — AI Investment Analysis Agent

FinAgent is an autonomous AI investment analysis agent built for VC/PE firms, fund managers, and analysts who need to evaluate companies quickly without a dedicated research team. The agent accepts two inputs: a financial report PDF and a stock ticker symbol. Once submitted, it automatically orchestrates a complete investment analysis workflow without human intervention. Core capabilities: Document Intelligence: Gemini's multimodal API reads annual reports and ESG documents natively, extracting key financial metrics, investment highlights, strategic risks, and ESG ratings directly from text, tables, and charts. Real-Time Market Analysis: Live stock data via yfinance including current price, 52-week range, moving averages (MA20/50/200), and technical trend signals. Valuation Models: PE, PB, PEG, EV/EBITDA, dividend model, and a real DCF calculator where users input growth rate and discount rate assumptions. Results are cross-referenced to produce an overall valuation verdict with confidence level. Peer Comparison: The agent automatically identifies 3-5 industry competitors and fetches their real-time market data for side-by-side comparison against industry and market averages. Cross Analysis: The agent synthesizes divergence signals between reported fundamentals and current stock performance, producing a one-sentence investment verdict. Enterprise Security via Lobster Trap: All LLM calls are routed through Lobster Trap, a deep prompt inspection proxy that detects prompt injection, PII exposure, and data exfiltration attempts in real time. A live security dashboard shows every AI interaction and its risk assessment. Three-tier output structure gives users a 30-second summary card for quick decisions, core analysis for standard review, and full detailed output for deep-dive research. All sections can be exported as a PDF report. Built on: Gemini API, React, TypeScript, Express, Vultr Cloud Compute, Lobster Trap (Veea).