Inspiration
Product Managers in large organizations deal with endless manual work: strategy documents, research synthesis, sprint planning, cross-team coordination, and vendor onboarding. At the same time, onboarding in regulated environments is slow, inconsistent, and expensive.
We wanted a single platform that solves both problems: an intelligent onboarding system paired with a productivity suite for Product Managers.
That became Syntria.
What it does
Syntria unifies two major capabilities into one platform:
- Intelligent Onboarding and Risk Analysis
AI-powered risk scoring using structured output
Detection of missing controls, PII exposure, and compliance gaps
Automated routing based on severity levels
Full audit trail for oversight and transparency
- Product Manager Productivity Suite
Strategy Agent generates PRDs, roadmaps, and user stories
Research Agent synthesizes customer feedback, competitor data, and market trends
Automation Agent creates reminders, tasks, and planning updates
Voice Agent allows hands-free interaction through ElevenLabs integration
- Unified Dashboard
One interface that brings together onboarding workflows, risk dashboards, PM tools, and audit trails for streamlined decision-making.
How we built it
Frontend: Vite, React, ShadCN UI components
Backend: Node/TypeScript with serverless API routes
AI: Gemini or OpenAI for structured, deterministic responses
Voice: ElevenLabs API
State management: React hooks and lightweight utilities
Design: Dark theme with gold and blue accents for a modern enterprise look
The system uses a multi-agent architecture, each designed with specific roles and JSON schemas to remain predictable and reliable.
Challenges we ran into
Stabilizing AI structured outputs under time pressure
Combining two different problem domains into a single, simple UX
Implementing serverless routes cleanly while managing environment variables
Keeping scope tight enough to finish while still delivering a high-impact demo
Accomplishments we’re proud of
A fully working onboarding pipeline using real AI risk scoring
Multiple AI agents that genuinely improve PM workflow
A clean, intuitive UI that judges can understand immediately
A solid architecture that runs locally and deploys easily
A feature set that feels practical and not just “for the hackathon”
What we learned
Well-structured AI outputs matter more than raw model power
Simplicity wins when combining complex ideas
Frontend clarity is just as important as backend capability
Working with multiple APIs under time pressure requires strict organization
Good project scoping is the difference between finishing and failing
What's next for Syntria
Full compliance reporting and automated document generation
Vendor knowledge graph with risk propagation
Memory-enabled multi-agent collaboration
Browser extension for PM workstreams
Notification system for real-time risk and workflow alerts
Enterprise role-based access control
Optional API so companies can embed Syntria onboarding into their own systems
Built With
- api
- apis
- elevenlabs
- gemini
- github
- javascript
- node.js
- openai
- react
- shadcn
- tailwindcss
- typescript
- ui
- vite

Log in or sign up for Devpost to join the conversation.