Inspiration
Startup founders and builders work at an unsustainable pace. We glorify "hustle culture" while ignoring the mental health crisis in tech. After watching countless talented founders burn out—not because their ideas failed, but because they ran themselves into the ground—I knew something had to change.
The problem isn't that builders don't care about wellness. It's that wellness tracking happens outside their workflow. By the time you remember to check in, you're already burned out. Builder Uptime solves this by integrating wellness monitoring directly into where builders spend their time: their browser.
What it does
Builder Uptime is an AI-powered burnout prevention platform that:
- Monitors wellness in real-time through a Chrome extension that tracks energy levels and focus quality throughout your workday
- Detects burnout risk using smart algorithms that analyze patterns and trigger interventions before you hit the wall
- Provides privacy-first mental health coaching via NEAR Protocol's Shade Agents—AI companions secured in Trusted Execution Environments (TEEs) so your mental health data never leaves encrypted space
- Integrates with your workflow through task management and productivity tracking, making wellness part of building, not separate from it
The platform catches you when your energy dips, reminds you to take breaks, and provides empathetic support—all while keeping your data completely private.
How we built it
Tech Stack:
- Frontend: Chrome Extension (Manifest V3) with React for the popup interface
- Backend: Node.js/Express API deployed on Railway
- AI/Privacy: NEAR Protocol Shade Agents (TEE-secured AI) for confidential mental health conversations
- Observability: Opik for LLM tracking and evaluation
- Real-time tracking: Browser-based wellness monitoring with in-memory storage
Architecture: The Chrome extension captures wellness data (energy levels, focus quality) and sends it to our backend API. When users need support, they chat with a Shade Agent—an AI assistant running in a secure enclave that ensures complete privacy. Opik tracks all interactions for quality assurance and continuous improvement.
Challenges we ran into
1. Opik SDK Integration Issues
The Opik npm package had dependency conflicts that crashed our Railway deployment. Solution: Built a custom REST API client that calls Opik's endpoints directly using Node.js's native https module—no SDK needed!
2. Privacy vs. Functionality Trade-off How do you provide personalized AI coaching while ensuring complete data privacy? Solution: NEAR Protocol's Shade Agents run in TEEs (Trusted Execution Environments), meaning AI processes sensitive mental health data in encrypted enclaves that even we can't access.
3. Chrome Extension Permissions Balancing functionality with user trust. We minimized permissions to only what's essential: active tab access for context-aware wellness checks, storage for user preferences.
4. Defining "Burnout Risk" Mental health isn't binary. We built a three-tier system (LOW/MEDIUM/HIGH) based on energy level patterns, with contextual recommendations rather than alarmist warnings.
Accomplishments that we're proud of
✅ Privacy-first AI wellness coaching using cutting-edge TEE technology
✅ Zero-dependency Opik integration that works reliably in production
✅ Real-time burnout detection that catches issues before they escalate
✅ Actually deployed and functional—live API with full observability
✅ Built for builders—integrated into workflow, not separate from it
What we learned
Technical:
- REST APIs > SDKs for production reliability when dealing with emerging tools
- TEEs (Trusted Execution Environments) are the future of private AI
- Observability is crucial for AI applications—you can't improve what you don't measure
Product:
- Wellness tools fail when they're separate from workflow—they must be seamlessly integrated
- Users value privacy over features when it comes to mental health data
- Small interventions (like break reminders) compound into major wellbeing improvements
Startup:
- Developer tools need exceptional documentation and clear error messages
- Railway makes deployment ridiculously easy once you get past initial config
- The hardest problems are often about human behavior, not technology
What's next for BuilderUptime
🚀 Integrate with productivity tools (Linear, Notion, Slack) for deeper context
🚀 Expand Shade Agent capabilities with personalized burnout prediction models
🚀 Build mobile app for 24/7 wellness monitoring beyond just work hours
🚀 Create founder community features for peer support and accountability
🚀 Partner with startup accelerators to make wellness part of the building process
🚀 Scale Opik evaluation with comprehensive LLM-as-Judge metrics for AI quality
Long-term vision: Make burnout as obsolete as paper filing cabinets. Every builder deserves to build sustainably—their mental health is their most valuable asset.
Try it: https://opik-liard.vercel.app/
Built With
- openai
- opik
- railway
- typescript
- vercel
Log in or sign up for Devpost to join the conversation.