StockIQ

A React client and an Express server. The server pulls live prices (GSE and NGX straight from the exchanges' own public market-data endpoints), merges them with curated GSE fundamentals, and streams a Gemini-written analysis to the browser over server-sent events. Results are cached so repeat requests stay fast and cheap.

• React
• Vite
• Tailwind CSS
• Node.js
• Express
• Google Gemini API
• Server-Sent Events
• Vercel
• GitHub Actions

Free market-data APIs such as Yahoo Finance don't cover the Ghana Stock Exchange or the Nigerian Exchange, so retail investors there have little structured research tooling.

• ▹Reliable data for markets that mainstream APIs ignore
• ▹Keeping AI output tied to supplied data instead of free-form opinion

• ▹Market dashboard and signal scan per market (GSE, NGX, JSE, US)
• ▹AI stock screener over a curated list of listed companies
• ▹Per-stock analysis grounded in live price data and GSE fundamentals
• ▹Streaming responses so analysis appears as it is written
• ▹Latest-filing checks for GSE annual reports

• ▹Use the exchanges' own public endpoints for GSE and NGX prices, since mainstream APIs don't cover them
• ▹Stream over server-sent events instead of waiting for a full response
• ▹Keep the Gemini model configurable with fallbacks so a model change needs no code change
• ▹Cache by response type with different lifetimes to limit LLM cost and latency

Content needed: measured results with period and method. Never estimated.

Evidence

• ▹Live frontend and API (the API's /api/health responds)
• ▹Public repository with client, server and deploy workflow

Content needed: what you would do differently

• AI Engineering · Full-Stack · Geospatial

  NAIMOS Intelligence

• AI Engineering · Applied ML

  Ghana Image Captioning

• Full-Stack · AI-Assisted Development

  LinkinForLife Business Manager