AI Engineer / İstanbul, Türkiye
I build production computer vision, ML serving and applied LLM systems, from the model and infrastructure to the product people actually use.
Featured production work
System views and product visuals from production and launch-ready work. Each project combines AI system design, backend engineering and operational ownership.
A production real-estate due-diligence platform that turns official cadastral, zoning, risk, mobility and demographic data into source-traceable parcel reports.
PostGIS spatial analysis, evidence maps, RAG synthesis, OCR and asynchronous data agents across a FastAPI, Celery, PostgreSQL and Qdrant stack.
A production API co-built by Ahmet Özel and Ahmet Güneş that turns PDF, DOCX, CSV, XLSX and ZIP files into retrieval-ready chunks and multilingual embeddings.
Sentence-aware chunking, BGE-M3 embeddings and a clean API for real-world RAG pipelines.
A closed-source ANPR and parking management product built around license plate detection, OCR and operational workflows.
A complete computer vision product with up to 99.9% recognition in optimal conditions, sub-600 ms processing and fully offline deployment.
A codebase assistant that answers questions across GitHub and Bitbucket repositories with grounded file citations.
RAG, context engineering and source-aware answers for understanding unfamiliar repositories faster.
Selected engineering projects
All repositoriesExperience
More than four years building AI systems at SBM, from an internship into a full-time engineering role, alongside founder-level product ownership.
Insurance Information and Monitoring Center (SBM)
Building production AI and computer vision systems for insurance workflows, with responsibility spanning models, optimized inference and the services around them.
Python / Computer Vision / ONNX / TensorRT / TritonBuilding AI products and delivery capabilities through Valnox, including EmlaxAI, PlakaNet and developer tools, while co-building Chunking Service.
Product Engineering / Geospatial AI / Backend / CloudExpertise
I work where models, infrastructure and products meet. That range makes the tradeoffs clearer, and the systems stronger.
Detection, OCR/ANPR and optimized inference for production and edge environments.
OpenCV / Detectron2 / MMDetection / YOLO / ONNX / TensorRT / TritonReliable training, serving and deployment paths with an eye on latency, scale and resource efficiency.
PyTorch / Docker / CI/CD / AWS / GCP / Model ServingRetrieval, embeddings, evaluation and agentic systems designed around real product constraints.
RAG / MCP / LangGraph / Hugging Face / RAGAS / LangfuseAPIs, asynchronous workflows and data infrastructure that keep AI products useful in production.
Python / FastAPI / PostgreSQL / MongoDB / Redis / SQLEducation
TÜRTEP / Remote
Maltepe University / English program, full scholarship
Contact
I'm open to AI engineering roles with remote work or relocation support.