Fintech Analytics Platform · Est. 2024
Markets,
quantified.
Live market data, ML forecasting, and an AI-assisted dev workflow across housing, equities, and crypto — all running on a single AWS stack.
Six surfaces, one stack.
Each section is a standalone dashboard drawing from the same Postgres database and Celery ETL pipeline.
Mortgage rates, ZHVI, and regional momentum.
FRED mortgage rates, Zillow ZHVI, median prices, regional heatmaps. 30-day LSTM forecast horizon.
Equities, treasuries, and signal detection.
OHLCV via Alpha Vantage, treasury yield curve from FRED, earnings signals, autoencoder anomalies.
Order flow, liquidations, and on-chain.
Binance WebSocket order flow, liquidation heatmaps, Fear & Greed, on-chain metrics.
Roadmap, deploys, and release health.
Three-DEV pipeline → QA → UAT → blue/green PROD. Gantt roadmap, sprint capacity, ADR log.
Service topology & runtime health.
Django + Next.js behind NGINX on a single t3.medium. Celery from RabbitMQ; Redis cache; Postgres truth.
RAG brain · structured prompts · agent chains.
All agent tasks are JSON-structured prompts enforced at the boundary. RAG-hydrated context; eval as first-class.