Senior Applied ML and GenAI engineer focused on measurable business outcomes. I build production AI systems that improve reliability, reduce operational risk, and accelerate delivery from prototype to stable release.

• 99.9% uptime supporting 100k+ daily API requests in production.
• 230% pipeline acceleration via workflow hardening and worker automation.
• 35% entity extraction F1 improvement in custom NER and JSON extraction systems.
• 98.7% OCR accuracy across 5M+ lines of low-quality legacy artifacts.
• 99.8% CI/CD reliability across 20+ AI automation releases.

LKGE (Lorekeeper Graph Engine): agentic story generation with Neo4j knowledge graphs, dual RAG retrieval, and contradiction guardrails.

• 61% fewer contradictions in paired evaluation versus rolling-context baseline
• Graph memory and pre-generation guardrails to preserve narrative consistency at scale
• Production-ready stack design with FastAPI, Streamlit, LangGraph, ChromaDB, and OpenTelemetry

Repo: https://github.com/darthmanwe/lorekeeper

Neo4j + FastAPI graph-backed retrieval and grounded RAG for medical documents.

• 5-stage Cypher retrieval pipeline with semantic reranking
• Sub-200 ms retrieval target
• Focus on provenance, explainability, and contradiction reduction

Repo: https://github.com/darthmanwe/Medical_Doc_Knowledge_Graph_System

Fault-tolerant distributed RL platform using Ray + PyTorch PPO.

• Asynchronous rollout collection across heterogeneous workers
• Heartbeat-based health monitoring
• Automatic worker replacement for resilience

Repo: https://github.com/darthmanwe/Training_Distributed_Systems

Document automation pipeline for converting instructional PDFs into structured, reusable presentation output.

Repo: https://github.com/darthmanwe/PDF_to_Presentation

Representative applied data science and ML implementation samples.

Repo: https://github.com/darthmanwe/Work_Sample

Prototype MCP-focused project with LEPOR evaluation concepts for AI workflow experimentation.

Repo: https://github.com/darthmanwe/Beats_MCP

• LLM / Agentic AI: LangChain, LangGraph orchestration, structured outputs, tool-calling workflows, context engineering, contradiction guard patterns
• Graph / Retrieval Systems: Neo4j 5.x, Cypher query design, Graph RAG + Vector RAG, ChromaDB, causal/event memory modeling
• ML / NLP / OCR: PyTorch, Transformers, OpenCV OCR, NER, extraction pipelines, evaluation and regression harnesses
• Backend / App Layer: Python, FastAPI, Flask, Streamlit, REST API design, Pydantic schema validation
• MLOps / Observability: Kubernetes, Docker, CI/CD validation environments, OpenTelemetry, Prometheus, pytest-based test suites
• Cloud / Data Platforms: AWS, Azure ML Studio, Databricks, GCP, Azure GovCloud, PostgreSQL, Supabase, BigQuery, Redis, Elasticsearch

• AWS Certified Machine Learning - Specialty
• Google Cloud Machine Learning Engineer
• Azure AI Engineer Associate

If you are hiring for senior AI/ML roles (GenAI, NLP/OCR, MLOps, production platforms), I am open to discussing high-impact opportunities.

• LinkedIn: https://www.linkedin.com/in/kutlu-mizrak/
• Email: mailto:[email protected]