Purwa Hartono

Purwa Hartono AI Engineer | LLM · RAG · Agentic (LangGraph multi-agent) | I build agents that cite every claim | Ex-CTO / Head of Data, now hands-on | MLOps + LLM eval | Remote (US/EU/SG/AU) Tangerang, Banten, Indonesia (ID) 490 connections • 514 followers About AI Engineer | LLM · RAG · Agentic. I build multi-agent and retrieval-augmented generation (RAG) systems to production standards with LangGraph, LangChain, and Python — evaluation, guardrails, and observability baked in. The path here wasn't a straight line: I started my career in 2015 as a full-stack developer, building an ERP for Honda IKM Group on.NET and Azure. A year later I stepped in front of a classroom as a programming lecturer — and discovered that building people scales even further than building software. In 2018 I built and led Purwadhika's Data Science & Machine Learning program — the school's best-selling bootcamp — and taught both data science and full-stack web & mobile development. Since 2016 I've taught 100+ engineers and data scientists hands-on, and the programs I built have produced 1,000+ graduates — now Data Scientists, ML Engineers, Data Engineers and BI/Data Analysts at leading tech companies, from Indonesia and Southeast Asia to the US and beyond. By 2022 I was Head of Data and CTO, leading the engineering, UI/UX, and data teams company-wide. In 2025, as AI Engineering Program Manager, I architected Purwadhika's 15-week AI Engineering program — machine learning & deep learning, computer vision, and NLP through LLM apps (OpenAI, Hugging Face), agentic AI, RAG, LLM observability (Langfuse), and MLOps (MLflow, Docker, GCP). In early 2026, after nearly a decade building and leading these programs, I made a deliberate choice: step out of leadership and return to hands-on engineering full-time. I'd rather build the systems than own the roadmap. Since then: three LLM systems, all with live demos. The flagship is a LangGraph multi-agent research assistant (Planner → Researcher → W