AI products designed to behave predictably in real world use.
Work centers on shaping raw model output into stable, usable features through clear constraints, thoughtful system design, and practical UX.
Reliability, clarity, and trust matter more than novelty. The goal is building AI tools people can actually depend on.
Controlled text transformation pipelines and guided model behavior, presented as a focused demo.
A compact rewriting and summarization tool built with the OpenAI API.
Demonstrates structured transformation pipelines, guided model behavior, and straightforward UX supporting predictable output.
Repo: https://github.com/Dhudson-AI/rewrite-studio
Submish (demo)
Multi step AI workflow centered on input validation, behavior control, and user friendly constraints.
Frontend and LLM demo featuring tone selection, clean input handling, and predictable output shaping. Emphasizes UI clarity, form validation, and safe request patterns.
Highlights UI clarity, form validation, and safe request patterns.