Metehan Berker
I'm a Machine Learning Engineer with 1 year of AI/ML experience and 2+ years of software engineering background.
About
I’m a Machine Learning Engineer with 1 year of AI/ML experience and 2+ years of software engineering background. I enjoy building multi-agent LLM systems, automating complex workflows with Python and modern AI frameworks, and delivering end-to-end ML pipelines from prototype to production.
I’m passionate about developing intelligent systems that solve real operational problems for businesses across diverse industries, and I thrive in cross-functional teams delivering practical, scalable solutions.
- 3 years' experience
- Joined March 2026
Skills· 16
Languages
- EnglishAdvanced
- JavaneseAdvanced
Industry experience
- Financial Services
- Manufacturing
- Software & Internet
- Professional Services
- Energy & Utilities
More experts in Los Angeles, United States
- AI Developers for hire in Los Angeles, United States
- AI Engineers for hire in Los Angeles, United States
- Back-End Developers for hire in Los Angeles, United States
- Data Scientists for hire in Los Angeles, United States
- Full Stack Developers for hire in Los Angeles, United States
- Web Developers for hire in Los Angeles, United States
Work experience
Architected a multi-agent Financial Wellness Advisor using LangChain and AutoGen, orchestrating seven specialized agents for budgeting, savings, and investment guidance with sub-2s response times. Implemented a RAG system with Pinecone vector database and OpenAI embeddings, improving advisor accuracy by 30% with Prometheus-tracked quality metrics and FastAPI REST endpoints. Engineered production-grade ETL pipelines using Apache Kafka and Spark to stream transaction data with sub-200ms latency, maintaining GDPR compliance and Zero Trust architecture throughout.
Trained and deployed object detection and regression models for monitoring marine power cables in floating offshore wind farms, as part of a Horizon 2020 project. Architected federated learning infrastructure using PyTorch time-series models across 5 distributed manufacturing sites, achieving 92% prediction accuracy while preserving data privacy. Built end-to-end MLOps pipeline with GitHub Actions CI/CD and Kubernetes, orchestrated inference, cutting model iteration cycles from 2 days to 6 hours with 94% deployment uptime.
Engineered high-performance C++ OMS and Compliance tools serving 200+ client portfolios and 1,000+ concurrent users with sub-millisecond latency and sub-second refresh rates. Delivered multi-threaded automated compliance validation system in C++ and Java, eliminating 90% of manual compliance checks and ensuring 100% regulatory adherence across 200+ portfolios. Re-engineered high-performance feed handler in C++ migrating from XML to JSON with FAST/FIX protocols, achieving 10x execution speed and 15x cache reduction through zero-copy deserialization.
Education· 6
Hire Metehan Berker today
Post a job and invite Metehan, or contact them directly. No hidden commission.










