SKIP TO MAIN CONTENT
TECHNICALInternship Sprint

IdeaEvaluator MVP

Monorepo delivering AI-assisted idea validation with a Fastify backend, Vite/Tailwind frontend, and Airtable telemetry integrations.

Role

Product Engineer

Category

Product Discovery

Skills

TypeScriptReactTailwind CSSFastifyPrompt EngineeringAirtable APINetlifyRender

Tools

ViteFastifyOpenAIAirtableNetlifyRenderVitest

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.