I'm Yohanes Andre Setiawan, an AI/ML engineer with four years of experience across education, e-commerce, and mining. I build production systems for LLMs and agentic AI, RAG, NLP, Information Retrieval, and RecSys, and I ship end-to-end ML pipelines from training to deployment. I recently completed my M.S. in Computer Science (AI) at Georgia Institute of Technology and currently work as a Senior AI Engineer at Datamine.

News

  • May 2026 🎓 Graduated with my M.S. in Computer Science (AI) from Georgia Tech. Continuing on as a Special/Non-Degree Seeking Student to keep doing research.

Experience

Datamine — Senior AI Engineer

Jan 2025 – Present

  • Engineer an LLM-powered RAG system (vector search + retrieval) for semantic search over product docs to replace keyword search.
  • Develop an MCP-based agent to automate multi-step workflows and reduce manual effort.
  • Design a multi-agent system that generates proprietary code to accelerate the SDLC.
  • Architect Azure CI/CD and MLOps for deployment/inference of AI services in production.
  • Integrate AI services into C#/.NET and REST APIs for production use.

Tokopedia — Data Scientist

Oct 2023 – Dec 2024

  • Implemented a two-tower retrieval/ranking model to improve recommendation.
  • Engineered product deduplication and feed diversification using clustering algorithms.
  • Integrated user negative-feedback into the recommender to reduce irrelevant results.
  • Improved age prediction accuracy by 5% using ensemble techniques, with an 80% reduction in feature reliance.
  • Built Airflow DAGs orchestrating C++ components for training and deployment.

CoLearn — Data Scientist

Sep 2021 – Oct 2023

  • Designed a recommender system using PyTorch transformers and FastAPI for personalized video delivery.
  • Built PoCs with OCR, NLP, and LLM prompting for a student teaching assistant.
  • Created GenAI/LLM solutions (RAG/retrieval, evaluation) using LangChain and Hugging Face.

Education

Georgia Institute of Technology

Special/Non-Degree Seeking Student

May 2026 – Present

Continuing research after my M.S. Coursework: Modern Internet Research Methods, Introduction to Research Seminar.

M.S. in Computer Science (AI)

Jan 2024 – May 2026

Completed the non-thesis track with a GPA of 3.90 / 4.00. Coursework: Machine Learning, Deep Learning, Knowledge-Based AI, AI for Robotics, Game AI, ML for Trading, GPU Hardware & Software, High Performance Computer Architecture, Network Science, Software Development Process.

Udayana University

B.S. in Electrical Engineering (Robotics)

Jun 2018 – Oct 2022

Head of the Robotics Club, managing 100+ members for the Indonesia Robot Contest. Graduated with a GPA of 3.88 / 4.00, culminating in a thesis on Computer Vision and Neural Networks in Robotics.

Projects

indocolbert

Late-interaction retrieval for Indonesian. Comparing ColBERT, dense, sparse, and hybrid baselines on MIRACL-id and TyDi QA-id.

grab-rag

Evaluating how small LLMs handle misleading retrieved evidence in RAG pipelines.

Open for Collaboration

I'm actively looking for PhD positions starting Fall 2027, focused on information retrieval, RAG, recommendation systems, and agentic AI. I'm also open to research collaborations in these areas. If you're a PI or researcher working on related problems, send me an email at [email protected].