Technical Lead & Senior AI Engineer. I build Agentic AI systems that go to production — multi-agent orchestration, RAG, LLM fine-tuning, MLOps. 5 years @ Litmus7.
I'm Jose Atlin — a Technical Lead and Senior AI Engineer with 5 years building production AI systems at Litmus7. My work lives at the intersection of Agentic AI, multi-agent orchestration, and MLOps. Not research, not demos — systems that handle real production load: an SRE automation platform that resolves incidents in under 5 minutes, a hybrid RAG pipeline with >75% retrieval accuracy across 50+ SOPs, LLMs fine-tuned with QLoRA and served via TGI, and a company-wide ML platform adopted by 4 engineering teams. I write about the unglamorous but high-impact side of AI engineering — the architecture decisions, tooling choices, and production war stories that don't make it into research papers. What you'll find here: Multi-agent system design and MCP integrations RAG pipelines that actually work in production LLM fine-tuning with QLoRA, PEFT, and HuggingFace TGI Python tooling and MLOps Lessons from shipping AI at scale Currently leading a team of 4 AI engineers. Open to senior AI Engineering and Agentic AI roles.
Markdown supported · ⌘↵ to post
bibin antony
Great read! I’ve been transitioning my own workflow to the Astral ecosystem recently, and the speed difference alone is a complete game changer. It’s so refreshing to move away from the 'dependency hell' and slow linting times of the traditional setup. Thanks for sharing your experience!