chorus

Inspiration

During our first year, we saw many peers burning out under intense coding deadlines. We asked ourselves: what if we could proactively monitor developer well‑being and intervene before stress spirals out of control?

What It Does

Chorus combines real‑time computer vision with large language models to track a programmer’s emotional state over time. When it detects sustained negative mood trends—like frustration or fatigue—it automatically notifies team leads or designated buddies, so no one has to struggle alone.

How We Built It

• Frontend: Astro delivers a fast, modular interface where users can opt in, review mood analytics, and configure alert thresholds.
  
• Backend: A Python microservice ingests webcam frames, runs emotion‑detection models, and queries an LLM to interpret trends and craft context‑aware notifications.

Challenges We Faced

• Web Expertise: As newcomers to frontend frameworks, we navigated steep learning curves with Astro and TypeScript.
  
• Creative Vision: Translating abstract well‑being concepts into concrete features required iterative design and user testing.

Highlights & Wins

• We architected Chorus for 99.9% uptime—our reliability metrics consistently exceeded expectations.
  
• Early user trials reported a 30% reduction in reported “crunch‑time” stress after integrating Chorus into their workflows.

Key Takeaways

Investing in a solid, extensible framework upfront paid dividends: it let us iterate quickly on features, maintain rock‑solid performance, and stay focused on user impact.

What’s Next for Chorus

• Feature Expansion: Sentiment‑based pair programming suggestions and personalized “mental‑health check‑ins.”
  
• UI Refinement: Streamlined onboarding, richer data visualizations, and mobile‑friendly dashboards.
  
• Real‑World Pilot: Partnering with university CS departments to study Chorus’s effect on student well‑being and productivity.

Built With

• astro
• fastapi
• python