# Thomas J. Vu

> Software engineer shipping agent systems, interactive products, and measurable AI infrastructure.

Thomas J. Vu builds AI companions, agent systems, and the animation and runtime glue that makes them feel alive. His software engineering has a game-dev brain: presence, memory, tools, and the loops that turn a model into something you can actually play with. Based in Boston, Massachusetts.

For agents: start with the homepage markdown, then the selected project pages. HTML pages also have `rel="alternate" type="text/markdown"` links. Contact: hello@thomasjvu.com. Resume: /thomas-vu-resume.pdf. LinkedIn: https://www.linkedin.com/in/thomasjvu. GitHub: https://github.com/thomasjvu.

## Start here

- [Homepage](https://thomasjvu.com/index.md): positioning, selected work, research, and contact
- [About](https://thomasjvu.com/about.md): bio and how he builds
- [Resume](/thomas-vu-resume.pdf): current PDF resume
- [Projects](https://thomasjvu.com/projects.md): full catalog
- [Full dump](https://thomasjvu.com/llms-full.txt): concatenated markdown for offline context

## Selected work

- [LLM Let's Play - Fire Emblem](https://thomasjvu.com/projects/llm-plays-fire-emblem.md): Agent gameplay harness: mGBA state in, tactical context to the model, controller actions back, dashboard for humans.
- [SteezUI](https://thomasjvu.com/projects/steez-ui.md): Shadcn-style design system with dual install paths, inspectable demos, and craft-focused primitives for design-engineering portfolios.
- [Rally](https://thomasjvu.com/projects/rally.md): Waifu visual novel chat experience with Live2D, memory, and a relationship loop you can actually play.
- [Mystery Gift](https://thomasjvu.com/projects/mystery-gift.md): TEE-backed verifiable randomness infrastructure for giveaways and gachapon systems without on-chain costs.
- [BS-Bench writeup](https://thomasjvu.com/writing/bs-bench-llm-deception.md): 600-game LLM deception benchmark

## Range

- [Alkahest](https://thomasjvu.com/projects/alkahest.md): Private and uncensored LLM inference platform with end-to-end encryption and trusted execution environment deployment.
- [Lisper](https://thomasjvu.com/projects/lisper.md): Gemma 4 audio speech-practice app for low-pressure lisp feedback, with public model artifacts and a browser-ready ONNX/WebGPU path.
- [Mystery Gift](https://thomasjvu.com/projects/mystery-gift.md): TEE-backed verifiable randomness infrastructure for giveaways and gachapon systems without on-chain costs.
- [Cloud Step 9](https://thomasjvu.com/projects/cloud-step-9.md): A browser-based rhythm platformer where you stomp on kawaii clouds to the beat of procedurally generated EDM.

## Writing

- [Lisper: AI Speech Therapy for Lisps](https://thomasjvu.com/writing/lisper-ai-speech-therapy-for-lisps.md): A quick writeup on Lisper, my Gemma 4 Good Hackathon submission for low-anxiety lisp speech practice.
- [BS-Bench: What a Bluffing Card Game Revealed About LLM Deception](https://thomasjvu.com/writing/bs-bench-llm-deception.md): A 600-game study of six language models playing Bullshit found that honesty prompts reduce deception, but also soften enforcement around the lies that remain.
- [Hardware Over Oracles: $0.01 TEE-Backed Randomness API (Offchain)](https://thomasjvu.com/writing/hardware-over-oracles-tee-randomness-api.md): Chainlink VRF is overkill for gacha games. We built a $0.01 alternative using Intel TDX.

## Optional

- [Garden notes](https://thomasjvu.com/writing): shorter definitions; skip unless the question is about a specific term
- [RSS](https://thomasjvu.com/rss.xml): writing feed