Working on
- Growth. Since August, helping grow sixonefour labs, an AI-native product engineering residency in Ho Chi Minh City, as it recruits its next cohort of engineers.
- This website. Lately: project case studies, topic pages for the blog, and a round of clean-up. The story of the rebuild is in the portfolio case study.
Learning and exploring
Most of my reading and experiments are about making AI agents dependable enough to ship:
- Spec-driven development and why requirements written clearly beat clever prompts. Start with Kiro and spec-driven development.
- Giving agents real tools safely through native MCP integration.
- Retrieval beyond the basics, and what changes when it runs in production.
- Securing LLMs and autonomous agents before they get access to anything important.
What's next
Near term:
- AWS certifications, to back up the hands-on cloud work with something verifiable.
- Projects. A few are half-built and a few are still ideas. The ones that get far enough will turn into case studies.
- Writing. More posts on the topics above, as I understand them well enough to explain them.
- Language practice, on a regular schedule.
- Closing my own gaps. Being honest about which skills are still thin, then picking projects that force me to learn them.
Longer term, I want to become a CTO and, along the way, a seasoned founder. That means learning beyond engineering too: the business, people and decision-making knowledge that a founder or technical leader needs.
Recently wrapped up
Roles that ended recently. The full list is on the experience timeline.
- AWS First Cloud AI Journey (April to July 2026): cloud engineer intern in the FCAJ bootcamp supported by AWS. The team project was DocuFlow AI, an invoice-processing pipeline on Step Functions, Textract and DynamoDB.
- VALSEA and GenAI Fund (May to July 2026): speech AI and real-time applications as an intern, and ambassador work for Agentic AI Build Week. The internship work is not public; the project I can show, built solo for the VALSEA EdTech Hack Sprint before I joined, is Classroom Copilot.
- TAIKHOANXIN and Luxe Wear AI (October 2025 to March 2026): two products built with small teams, documented as case studies: TAIKHOANXIN and Luxe Wear AI.
Say hello
Open to junior and fresher roles in cloud and AI engineering. If that, or anything above, overlaps with something you are building, email me at trahoangdev@gmail.com.