← All projects
AIFull-Stack
Pulse
Turns anonymous employee feedback into issues leaders can act on.
Pulse
The problem
Raw anonymous feedback is noisy and hard to act on. Leaders drown in comments and can't see the patterns underneath them.
The approach
Employees submit feedback by text or voice. AI clusters and synthesizes it into structured issue cards with severity and themes, shown on a leadership dashboard, with a keyword fallback when no API key is configured.
AI architecture
GPT clusters and summarizes submissions into actionable issue cards; Deepgram transcribes voice feedback. A privacy-first design keeps every submission anonymous.
Highlights
- ▹Anonymous-by-design submissions
- ▹AI clustering into actionable issue cards
- ▹Voice input via Deepgram transcription
- ▹Leadership analytics dashboard (Recharts)
- ▹Graceful offline keyword fallback