TT-Pulse is a high-performance monitoring ecosystem designed for local infrastructure, bridging the gap between raw system telemetry and developer productivity. Built with a "Security-First" mindset, it provides real-time observability into distributed nodes while ensuring absolute data privacy.
TT-Pulse transforms standard server monitoring into a proactive SRE (Site Reliability Engineering) tool. It treats home-lab nodes as "Active Workstations" and correlates system health with development velocity.
- Grinding vs. Sleeping: Real-time status tracking using high-precision heartbeats.
- Conquest Correlation: Tracks GitHub activity alongside CPU/RAM trends to identify how development sprints impact system load.
The ecosystem features Texas, a Bengal-inspired AI persona that acts as the studio guardian.
- Proactive Insights: Correlates telemetry data with Git commits to provide witty, actionable observations.
- Resilience Engineering: Powered by a custom SRE Fallback Engine. While architected to support Gemini 1.5 Flash, the system currently utilizes a high-performance heuristic engine in
dashboard/app/api/insights/route.tsto maintain 100% uptime during API rate limits or outages.
- Frontend: Next.js 16 (utilizing Turbopack for lightning-fast HMR) and Tailwind CSS 4.
- Visualization: Tremor for low-latency, time-series data rendering.
- Backend: Supabase with Real-time CDC (Change Data Capture) for instant UI updates.
- Edge Agent: Lightweight Node.js collectors running on Arch Linux, communicating via a secure Supabase backbone.
Data privacy is not an afterthought; it is baked into the core architecture:
- The Data Washer: A specialized sanitization layer in the agent (
ops.js) that strips sensitive information. It automatically removes absolute file paths (e.g.,/home/user/) and process arguments before they ever leave the local machine. - Defense-in-Depth: Implements Row Level Security (RLS) at the database layer, ensuring that telemetry data is strictly isolated and accessible only to authorized sessions.
The dashboard is designed for high availability. In the event of an upstream LLM outage, the SRE Fallback Engine takes over, ensuring the "Texas" persona remains active and providing insights based on local deterministic patterns.
Developed for TT Family's Dev Studio — Professional Development, Family Owned.