September 2024 – Present
Python Developer
9epoch.ai
Melbourne, Australia (Remote)
Production AI and data engineering for financial intelligence.
- Developed and operated a production multimodal LLM pipeline that converts broker emails, PDFs, and images into structured financial intelligence, including companies, recommendations, target-price changes, sentiment, and actionable alerts.
- Built and deployed an authenticated Model Context Protocol (MCP) server enabling AI clients to securely access financial databases and server-side resources through typed, validated tools.
- Enhanced an AI-generated Morning Note with ranked trade ideas, expected-move analytics, broker alerts, and controlled historical context for investment workflows.
- Improved LLM reliability and efficiency through structured outputs, deterministic validation, multimodal analysis, model routing, prompt caching, parallel inference, and regression testing.
- Built real-time ETL pipelines using Python, AWS Lambda, ECS, SQS, ElastiCache, PostgreSQL, and CDK to ingest and deliver financial data for downstream analytics.
- Backtested NLP-based broker signals against historical market-share data using statistical significance tests and risk-adjusted performance metrics.