hamza Khan

About

Backend/AI engineer who built production-grade intelligence systems in high-stakes domains including tax/legal document analysis and brain tumor MRI workflows. Stands out for combining LLM/RAG product delivery with strong engineering rigor around retrieval evaluation, grounding, validation, observability, and safe fallbacks—turning impressive demos into systems users could actually trust.

Key Strengths

• Owned backend systems end-to-end from design through launch and operations
• Built document intelligence backend for tax/legal workflows processing ~100,000 documents/day
• Improved retrieval quality with hybrid search, reaching ~89% retrieval accuracy
• Strong evaluation mindset using Recall@k and MRR instead of subjective output review
• Experienced in reliability engineering for high-stakes/compliance-heavy systems
• Designed async ingestion architectures to trade latency for reliability and scalability
• Strong grounding and validation approach to reduce hallucinations in LLM systems
• Incident ownership and prevention in regulated workflows
• Applied AI backend experience across legal/financial documents and medical imaging
• Clear separation of deterministic ML outputs from LLM-generated explanations

Experience

Education

Skills

Languages