About

About Brett Haas

I’m a software engineer in Bellevue, Washington, building ambitious products across AI, web, mobile, and intelligent systems.

Focus

My work sits where dependable product engineering meets applied AI research: shipping cross-platform mobile and cloud software, red-teaming and evaluating frontier language models, and studying how inference-time interventions change model behavior without retraining.

Education

B.S. in Computer Science, University of Virginia (Charlottesville, VA), May 2026. Focus areas: Computer Science, Cybersecurity, Machine Learning, and Software Engineering.

Experience

  • ML Research Assistant, University of Virginia, Nov 2025 – May 2026. Inference-time activation-steering research focused on language-model visual bias.
  • GenAI Technical Advisor Intern (SEAL), Scale AI, Jun 2025 – Dec 2025. Frontier-model safety, evaluation, and agent reliability work on the Scale AI SEAL team.
  • Software Engineer Intern, Refraction Innovation Hub, Jun 2025 – Aug 2025. Cross-platform nutrition and food-recognition product development across mobile, AI, and cloud systems.

Role-by-role highlights are on the experience page.

Selected projects

  • Court Vision: Computer vision basketball analytics platform for shot tracking and player movement insights.
  • Beat Stream: A real-time collaborative music production platform built for low-latency multi-user sessions.
  • Vision Bias Steering: LLM steering experiments for shifting model outputs between spatial and descriptive language.

Tools

  • Languages: TypeScript, JavaScript, Python, Java, C/C++, Go, Rust, SQL, Bash, HTML/CSS.
  • Frameworks: React, Next.js, Node.js, React Native, Flask, PyTorch, TensorFlow, NNSight, Claude Code.
  • Cloud & DevOps: AWS, Docker, Kubernetes, CI/CD, Google Cloud, Linux.
  • Databases: PostgreSQL, MySQL, MongoDB, Redis, Azure SQL.
  • AI / ML: OpenAI API, ML Data Pipelines, RLHF, LLM Evaluation, Activation Steering, Mechanistic Interpretability, Prompt Engineering.

Elsewhere

Email [email protected], browse code on GitHub, or connect on LinkedIn. How this site handles data is described in the privacy policy.