Building AI-powered platforms to automate and orchestrate workflows in private financial markets

Built enterprise-level AI tech for Investment Banks and Private Equity firms. Primary focus was on developing low-latency pipelines that could accurately recreate tables and charts. Additionally architectured and built an Excel Add-In for the ProSights platform.

• Boosted deployment efficiency by 41% and trimmed memory usage by architecting scalable back-end services on DigitalOcean using Docker and SQL, ensuring peak performance for the Ticketdude app now live on the Play Store.

• Streamlined API P90 latency by 29% through the implementation of a QR code-based check-in system with Python, Django, and Redis caching, significantly enhancing participant event entry efficiency.

• Accelerated server-side processing by 30% by developing custom registration forms with Python and Django, utilizing gRPC for enhanced inter-service communication.

• Expedited development time by crafting a cross-platform mobile app using Flutter and a progressive web app with React and TypeScript, focusing on front-end optimization, UX, and automated CI/CD pipelines via fastlane.

• Engineered a real-time notification system with Node.js, WebSocket & Google FCM and increased event interaction by 25% while integrating a payment gateway and job queue for improved transaction reliability.

• Elevated team productivity through leading the deployment of a Spotify Backstage instance, centralizing developer resources and streamlining service access for enhanced collaboration.

• Reduced search latency by 20% by architecting a scalable backend for an AI-powered medication search engine using Node.js, TypeScript, and PostgreSQL, integrating ChromaDB and Llama 3 to enhance RAG capabilities.

• Boosted data extraction efficiency by 30% and cut data matching errors by developing a self-healing web scraper using fine-tuned LLMs and embedding models, ensuring accurate medication data extraction.

• Trimmed API response time by 20ms by implementing an event-driven microservice architecture with AWS Lambda and Node.js, significantly improving API Requests per minute (RPM) for the prescription search engine.

• As a team, we architectured, built and deployed “Qubit”, an AI based coding assistant meant to introduce and teach coding to high school students

• Built and deployed an AI based language learning application that helped students understand Italian with near real time translations and transcriptions