Thanh Phong Trần

I’m an AI Engineer specializing in LLMs and AI Agents, with hands-on experience designing and developing LLM-powered applications and multi-agent workflows. I focus on RAG systems, prompt engineering, tool calling, agent orchestration, and guardrails to deliver reliable outputs.

In my current role at Vinsmart Future, I designed and developed an AI-powered learning material system, including an exam generation pipeline, a student learning assistant chatbot, and learning analytics using multimodal parsing, RAG, ReAct Agent patterns, and multi-agent prompting. I also built an enterprise AI Evaluation Platform to assess the quality of AI agents and LLM applications across products, leveraging agent evaluation, LLM-as-a-Judge metrics, and production monitoring.

Before that, I worked as a Quantitative Researcher at Finpros Investment Joint Stock Company, researching quantitative trading strategies, building backtesting frameworks, and supporting deployment of automated trading systems on cloud servers. Earlier, I completed a Data Engineering Internship at FPT Software, developing data processing pipelines with BigQuery and Composer (Airflow) and assisting with deploying machine learning models using BigQuery ML.