Salih Hasan S.

I build AI systems that actually run in production, not just demos. Most clients come to me after hitting the same issues: RAG not accurate, agents unreliable, or AI pipelines that don’t scale. That’s where I step in. I’m a Senior ML and LLM Engineer (10+ years), recently working at NexGen Cloud and Holi GmbH, where I built: ☑️Enterprise AI platforms serving 100+ clients with 99.9% uptime ☑️RAG systems (multimodal, hybrid search, real-time pipelines) ☑️AI agents connected to APIs, tools, and business workflows ☑️High-performance inference systems (sub-200ms latency, cost reduced ~40%) What I can help you with: ☑️RAG Systems (LLM + Data) PDFs, databases, internal tools → accurate retrieval (LangChain, LangGraph, pgvector, Pinecone) ☑️AI Agents & Automation Multi-step workflows, tool calling, OpenClaw-style systems ☑️LLM Systems (Production Ready) OpenAI, Claude, Gemini, structured outputs, evaluation, guardrails ☑️Backend + APIs FastAPI, Django, Node, async pipelines, Redis, scalable systems ☑️Search & Recommendations Hybrid search (BM25 + vector), ranking, personalization I’ve built systems handling 1000+ ML jobs/month and real traffic I focus on accuracy + reliability, not just “AI output” I design systems your team can actually maintain 𝐐𝐮𝐢𝐜𝐤 𝐰𝐢𝐧𝐬 𝐈’𝐯𝐞 𝐝𝐞𝐥𝐢𝐯𝐞𝐫𝐞𝐝 Improved search latency and relevance by 40% Increased recommendation quality by 35% Reduced AI infrastructure cost by up to 70% If you’re building something serious with AI and want it done properly, 𝐥𝐞𝐭’𝐬 𝐭𝐚𝐥𝐤.