AI/ML Engineer building production LLM & RAG systems
Selected Projects
Secure document search and AI chat across workspaces, with RBAC, JWT auth and tenant isolation. Async ingestion (OCR, chunking, embedding), hybrid retrieval (Qdrant + BM25 with RRF and reranking), cited answers, and Prometheus/OpenTelemetry observability with Recall@K, MRR and nDCG evaluation.
Full-stack AI video generation platform for consistent fitness videos from reusable characters, environments and prompt templates. Provider-agnostic layer over local ComfyUI and cloud models, async job orchestration, and a Next.js dashboard for assets, prompts and job tracking.
Ranks candidates against job descriptions with semantic embeddings and pgvector search. Hugging Face embeddings, OpenAI-powered job description enhancement, and background processing with Celery, Redis and Docker.
Freelance AI service that turns PDFs into bilingual (Persian/English) multiple-choice questions. Semantic chunking, duplicate detection, answer validation, retry and response repair, with unit-tested core components.
From research to production. I build LLM applications, retrieval pipelines and backend systems that are measurable, observable and ready to scale. Models are only useful once they ship.
I am a Backend Developer Turned AI/ML Engineer
I specialize in production LLM applications, Retrieval-Augmented Generation and scalable backend systems, from AI services and vector search to distributed ML infrastructure. I'm completing an M.Sc. in Soft Computing and Artificial Intelligence at the University of Tabriz on a full scholarship. Based in Tabriz, Iran, and open to relocation.
And This Is My Career
AI/ML Engineer at NetPardazAzar 09/2026 - Now
Building AI and machine learning solutions, applying my LLM, RAG and backend experience to production systems.
AI/ML Engineer at Desmer Guvenlik 02/2024 - 03/2026
Deployed AI optimization models for logistics across 100+ service points, cutting service delays by 30%. Used causal inference to evaluate process changes (20% fewer delivery delays), integrated ML into live systems (40% less manual oversight), and forecast ATM cash demand across 50+ locations, reducing cash shortages by 25%.
Backend Developer at Aral Studio 02/2023 - 01/2024
Designed real-time backend systems for high user traffic with excellent uptime. Improved API response time by 35% through refactoring and database indexing across 10+ endpoints, and architected microservices that cut deployment time by 40%.
Backend Developer at NOAY 09/2021 - 01/2023
Built secure, scalable backend infrastructure for a cryptocurrency platform handling high-frequency transactions with near-zero failures. Developed APIs and microservices that reduced server load by 25%, working closely with frontend teams.
Academic Background
M.Sc. Soft Computing & AI, University of Tabriz 2023 - Now
Selected as a "Brilliant Talent" with direct admission and a full scholarship. Grades: Artificial Intelligence A, Data Mining A+, Deep Learning B+, Machine Learning B.
B.Sc. Computer Science, University of Tabriz 2019 - 2023
Full scholarship, ranked 4th in overall GPA, top 3% in the national university entrance exam, and honorable mention at the ICPC in 2019, 2020 and 2021.
I'll help your AI project ship
End-to-End Pipelines
From ingestion, chunking and embeddings to hybrid retrieval, reranking and cited answers, I own the whole RAG stack.
Measurable Impact
Delivered 30% fewer service delays and 25% fewer cash shortages. I evaluate with Recall@K, MRR and nDCG, not vibes.
Production-Ready
Auth, tenant isolation, async workers, Docker and observability with Prometheus, OpenTelemetry and Grafana.
AI & LLM Engineering Course
A course on my YouTube channel, X OFF SCRIPT, from the basics of LLMs to building real AI systems.
Episode 01 AI Engineering — Course Introduction
AI EngineeringOff the Clock, On the Mic
Away from engineering, I host a podcast called 91.8 Radio. New episodes are shared only on my Telegram channel.
91.8 Radio
Relaxed conversations away from the keyboard. 2 episodes so far, with more on the way.
Let's Get in Touch
Have an LLM, RAG or backend project in mind? Let's talk.