LLM Engineer

I'm Burak, I'm 27, I'm an Industrial Engineering graduate from one of Turkey's top universities (2020). During university years, I built a YouTube channel with 15k+ subscribers on ML, published a Udemy course on image AI models, interned at Stanford's Biomedical Data Science faculty, co-authoring a preprint for healthcare AI models. At 2021, I was a data scientist at electric distribution company.I created an MLOps system to forecast electricity usage for 8 cities using Python, Postgres, weather APIs, and tree models like LightGBM/XGBoost, with MLflow for auto feature engineering, ensembling, and daily predictions. Between 2021 and 2024, I spent 3 years at Gesund as the first engineer and main contributor. It’s a healthcare MLOps startup that was listed in CB Insights’ AI 100 alongside OpenAI at 2024. I've been the main developer of MLOps platform that handles Tensorflow/Pytorch image models — handling medical data, on-prem model deployment using Docker, FastAPI, MongoDB, RabbitMQ, EC2, ELK(elastic stack) and S3. I developed detailed validation pipelines for classification, detection, and segmentation models, with fairness checks across age, gender, and ethnicity and other healthcare metadtas. I also created an AutoML module, an explainability library. Beyond engineering, I represented the company at conferences, joined meetings with FDA and White House reps, and managed software side of $400k+ contracts, and published a paper there. After leaving the Gesund, I was about to be cofounder of developed two LLM+RAG apps for startups—one a financial VC due diligence tool (backend with RAG, no Langchain), and another a YC application focused on growth teams. I joined Career.io in September 2024 as an ML Engineer, I'm fully owning the AI Interview Preparation Mocking App. It's an LLM and RAG product which having 5,000+ monthly active users, bringing relevant questions with vector databases and utilises agents to continue interview and give detailed feedbacks. I also worked on HR projects like training small language models for job classification for our job listing app that consumes 500k+ data weekly, resume parsing, tailoring, and review using ChatGPT APIs. On the engineering side, I moved old projects to Postgres/Argocd, managed CI/CD with GitHub Actions, I deployed via ArgoCD, tracked issues via Sentry,and used Chroma for retrieval projects.