Buse K.

I build and ship AI and data systems in production. Currently as a Data Scientist / AI Engineer in finance, working on machine learning, LLM applications, agentic systems, and the data pipelines underneath them. Alongside that, I've spent two years writing technical content professionally for companies in this space. I take on both kinds of work, and the overlap is the point: I write about systems I've built, and I build with an eye for what will actually need explaining later. Engineering * RAG and LLM applications. Retrieval design, vector databases, evaluation with RAGAS, tracing with Langfuse * Agentic systems built with LangGraph and smolagents * Data pipelines and monitoring. Snowflake, Airflow, dbt, streaming ingestion * ML workflows, NLP, OCR automation, and cloud deployment (Azure, AWS, Docker, FastAPI) Technical writing * Hands-on tutorials with working code, built and tested before publishing * Architecture explainers, use cases, and developer-focused long-form content * Editor at Data Science Collective, one of Medium's largest data science publications (950K followers) * Published technical work for Octoparse, Estuary, QA.tech. * A recent piece on real-time pipelines for RAG and fraud detection reached 18.4K views How I work Before writing about a product, I use it. I build something real with it, note where I got stuck or where the docs fell short, and put those observations in the work. The same instinct applies when I'm building: I document the friction while it's still fresh, not after.