# Giorgio Sarno

> Giorgio Sarno is an AI Software Engineer at Replika, working on LLM serving for a consumer app used by millions, agent infrastructure, and long-term memory.

Current scope: vLLM and SGLang serving decisions across latency, GPU cost, reliability, model quality, and capacity; agent tools, background tasks, and coordination across multiple agents.

Leadership and earlier work: Head of Data & AI at Sparrow Parts and AI Team Lead at Stratio; led teams of up to eight engineers. Designed and deployed embedding models for industrial part matching, and led a team that brought time series models into production for fleet monitoring.

Skills: LLM inference, vLLM, SGLang, KV caching, agents, agent memory, MCP, PyTorch, Kubernetes, CUDA.

PhD in physics (numerical work in C/CUDA).

## Profile

- [Giorgio Sarno](https://giorgiosarno.com/): current focus, work history, technical skills, and contact links.

## Writing

- [Writing index](https://giorgiosarno.com/writing): published articles on LLM serving, agent memory, and AI coding agents.

## Public profiles

- [LinkedIn](https://www.linkedin.com/in/giorgio-sarno/): professional profile and contact route.
- [GitHub](https://github.com/topppg): public code and profile README.
- [X](https://x.com/giorgiosrn): short notes on AI engineering.