Nadir Z.Private profile: Upwork shows its earnings only to signed-in members, so we check it every few weeks, not daily. MRR here is a rough estimate (Upwork's six-month earnings ÷ months active). Last read Aug 29, 2026.

I build production AI systems with LLMs, RAG pipelines, and voice agents - backend-first engineering on Python (FastAPI) and Node.js (NestJS), with full-stack delivery when you need it. Most AI projects break the moment real users hit them: hallucinations, slow retrieval, broken function calls, runaway token costs, voice agents that lose context. I fix that. I've shipped end-to-end AI systems across RAG pipelines, document processing, conversational chatbots, voice AI, multi-step agents, and fine-tuned models - all built to hold up in production, not just demo well. What I build: - RAG pipelines - embeddings, vector search (Pinecone, FAISS, pgvector), re-ranking, retrieval evaluation - LLM API integrations - OpenAI, Claude, structured JSON outputs, function calling, tool use, guardrails - Document processing - PDF parsing, OCR, structured extraction at scale - AI chatbots & conversational agents with memory, context management, and grounded responses - Voice AI systems - speech-to-text, TTS, real-time voice agents (Twilio, Vapi, LiveKit-style architectures) - AI agents - multi-step workflows, tool calling, planning, recovery from failure - Fine-tuning - custom model training on domain data (OpenAI, open-source) - Backend systems - FastAPI (Python) and NestJS (Node.js) with clean architecture, auth, queues, observability - Full-stack delivery - Next.js / React frontends when the project needs a working product, not just an API - AWS deployment - EC2, Lambda, S3, RDS, with cost and latency monitoring Recent results on Upwork: ✓ AI/ML Full-Stack project delivered in a week - 5-star review, client called the work "professional" and would hire again without hesitation ✓ Full-Stack contract over 25 hours - 5-star review for clean code, on-time delivery, and clear communication Tech stack: Python, FastAPI, Node.js, NestJS, Next.js, React, OpenAI, API, Anthropic Claude API, LangChain, LlamaIndex, Pinecone, FAISS, pgvector, PostgreSQL, Redis, AWS (EC2, Lambda, S3, RDS), Docker, Twilio, Vapi. If you're past the prototype stage and need an AI system that's reliable, accurate, and ready for paying users - message me with what you're building or where it's breaking. I'll reply with how I'd approach it and what the first milestone looks like.