Enrichyouenergy

AI Applications · 2026

ScottyLabs AI Agent Development

I improve Bark’s memory retrieval, tool discovery, and streaming performance.

The ScottyLabs dog outline and wordmark in white on a purple, blue, and pink gradient.

What the team does

I’m part of the ScottyLabs AI Agent team, working on Bark, an AI assistant for Carnegie Mellon students. My upcoming work focuses on:

  • Replacing keyword-based memory retrieval with context-aware retrieval.
  • Improving tool discovery beyond keyword-based loading.
  • Parallelizing input moderation with response generation.
  • Eliminating duplicate moderation checks.
  • Implementing incremental, non-blocking output safety checks to reduce streaming latency while preserving prompt-leak and secret protection.

Through this work, I’m learning backend development, API integration, testing, and collaborative software development. It’s also helping me understand how AI agents select tools, recall context, and balance responsiveness with moderation.