Hayk K.

I design and build production AI systems for SaaS products—document and CV parsing, RAG search, scoring, matching, and recommendation engines.

Unlike prototypes that work on a handful of examples, my systems are built for messy real-world data, predictable structured output, validation, monitoring, and controlled LLM costs.

RECENT WORK

• An AI recruitment platform processing approximately 5,000 CVs per day
• Semantic search and AI job matching for a leading recruitment platform
• Multi-provider LLM pipelines using OpenAI, Gemini, Qwen, and Ollama
• APIs, queues, concurrent workers, and event-driven backend services
• A production cloud migration that reduced infrastructure costs by 40–70%

HOW I WORK

I'm a hands-on technical lead: I design the architecture, write the critical code, review implementation, troubleshoot production systems, and ship. You get one person who understands the whole system instead of a handoff between an architect and an implementer.

CORE STACK

Python, Go, FastAPI, Node.js, PostgreSQL, Elasticsearch, vector databases, Docker, Kubernetes, AWS, and OVH.

15+ years in engineering | $500K+ earned | 31,000+ Upwork hours | 100% Job Success | Top Rated Plus

If your AI feature works in a demo but fails on real data or real volume, send me the architecture or pipeline. I'll give you a direct assessment of what should be fixed and what is worth building.