AI / ML Engineer · Full Stack Engineer · Open Source
AI / ML Engineer at SpazioDati (ION group), 4+ years building production LLM systems end-to-end — multi-agent platforms, RAG pipelines, NLP classification, and the inference infrastructure underneath, with reach from GPU-served stacks through Next.js frontends. Shipped Langfuse Freeze on PyPI and npm. M.Sc. Data Science · B.Sc. Philosophy.
llm-elasticsearch-cache ↗ — ElasticsearchCache + ElasticsearchStore: LLM completion cache and embedding vector store; now part of the official LangChain library.
langfuse-freeze ↗ · langfuse-freeze-js ↗ PyPI npm — offline-resilient wrapper for the Langfuse prompt SDK; snapshots prompts to disk at startup and falls back transparently when Langfuse is unreachable.


