AI Engineer building local-first RAG systems & evaluation tools

Hey everyone!

I'm Abdelatif, an AI Engineer based in France, working on Generative AI, RAG systems, LLM evaluation and AI agents.
I'm particularly interested in building reliable, privacy-first AI systems that can run locally or in on-premise environments.

What I'm building:

I'm the creator and maintainer of RAGnarok-AI, an open-source framework for local RAG evaluation.
The project focuses on making RAG systems measurable and reproducible, with evaluation metrics, local LLM judges, benchmarking tools and CI/CD quality gates.

I'm actively developing it and exploring ways to make LLM evaluation more reliable, especially when working without external APIs.

I've also worked on production-oriented GenAI applications, including document retrieval, agentic workflows and human-in-the-loop validation in privacy-sensitive environments.

What I'm interested in:

RAG evaluation and benchmarking.
Local LLMs and privacy-first AI architectures.
Agentic systems and practical LLM applications.
Open-source collaboration and real-world AI engineering.

One thing I'm looking for:

I'd love to connect with other builders working on RAG evaluation, local AI and LLM reliability. I'm particularly interested in exchanging ideas, getting feedback on RAGnarok-AI and exploring potential collaborations.

Always happy to connect, share experiences and learn from what others are building!