Tommy Fricano.
I'm a Solutions Consultant at Fullstory, where I design AI agent architectures and enterprise integrations for clients in financial services, tech, and consumer data. Before that, I spent two years architecting Azure systems for Circle-K's 200M+ daily operations. Outside of client work I'm building an autonomous referee for table tennis and a fully local voice assistant for my apartment — I like projects that force me to learn something new.
- Currently
- Fullstory
- Based in
- Atlanta, GA
- Studying
- M.S. AI · Georgia Tech
Recent projects
See all →Homy — Local Voice Assistant
A fully local LLM agent (zero external API calls) for the home, with tool-calling, layered memory, and a governed reviewer/worker pipeline for autonomous task execution, controllable by voice from a smartphone on the local network. Built a three-tier memory architecture — in-memory working context, summarized episodic recall, and RAG-based semantic memory using local embeddings with top-k search — giving the agent persistent memory without a hosted vector database.
Table Tennis Score Keeper
An autonomous table tennis scoring system that uses computer vision and machine learning to automatically track and score matches in real-time. The system integrates with a platform to record match history, track player statistics, and maintain a competitive leaderboard. Currently in active development.