Project Overview
IdeValidator helps founders stress-test early product ideas through multi-lens AI evaluations. The repo is structured as a npm workspaces monorepo containing a Vite + React frontend and a Fastify backend. Each analysis runs five cognitive "lenses" in parallel—Feasibility, Market Fit, Differentiation, Business Model, Visionary—and synthesises findings into actionable advice.
Sed me an e-mail if you want to try the system out.
Key Features
- Guided idea submission flow with validation, character budgeting, and keyboard shortcuts to trigger analysis.
- Results dashboard presenting per-lens rationale, pros/cons, and a composite gauge along with export/save actions.
- Airtable-backed telemetry capturing lens timing, token usage, and feature flags without storing raw ideas.
- Nimbus AI research integration that surfaces relevant reports to contextualise each evaluation.
Technical Highlights
- Frontend built with Vite, TypeScript, Tailwind, and context hooks for auth/session management.
- Backend Fastify service exposing
/analyze, logging with Pino, and delivering OpenAPI documentation from the monorepo. - Prompt schemas enforce deterministic JSON output, enabling strong typing across frontend and backend.
- Airtable automation scripts (
scripts/createAirtableSchema.ts,scripts/seedAirtableLogins.ts) bootstrap operations with reproducible infrastructure.
Results
- Achieved sub-2-minute end-to-end evaluations across five lenses while maintaining P95 backend latency under 10 seconds.
- Content and telemetry separation ensures GDPR alignment: only metadata hits Airtable while ideas remain transient.
- Modular design and REQ-st trace comments keep future features traceable and auditable.
Next Steps
- Layer in collaborative workspaces so multiple founders can iterate on the same idea history.
- Ship Vitest coverage for lens utilities and integrate Playwright smoke tests before production.
- Extend Nimbus research with live funding data and case studies for industry-specific insight.