Algorithmic Biography-Matching & Real-Time Market Calibration for First-Time Founders.
FoundersRadar is a global, mass-market B2C SaaS platform ($14.99/month) designed to eliminate the 90% failure rate of first-time startups due to "founder-to-idea mismatch." The platform analyzes an aspiring entrepreneur's unique career history (Person A) and cross-references it against a global vector database of historically successful entrepreneurs (Person B) to output a highly personalized, market-validated, month-by-month startup roadmap.
CRITICAL STRATEGIC FOCUS: We are currently in the Pre-Seed / Pre-Launch Validation phase. To protect company capital and safeguard intellectual property from competitors, DO NOT connect or deploy live third-party LLM APIs (OpenAI/Anthropic) to the frontend client.
This version of the codebase acts as a local experimental matching engine embedded into the existing company website's SERVICES section. It uses a curated sample dataset and does not claim to provide objective business validation.
[User Resume / Profile Text] ──► [Local UI Verification] ──► [On-Device PII Masking]
│
▼
[Pre-Seed Investor Metrics] ◄─── [Supabase Database] ◄─── [Secure Waitlist Data Ingestion]
- Frontend Framework: Next.js (React)
- Styling: Tailwind CSS (Mobile-First, Responsive Design Layouts)
- UI Components: shadcn/ui (Accessible HTML Primitives)
- Database & Auth: Supabase / PostgreSQL (
foundersradar_waitlistcollection) - Development Framework: Kiro IDE (Spec-Driven Development Environment)
- Service Hero Card: A modern, clean container introducing the platform value proposition with clear visibility of the target mass-market pricing ($14.99/mo).
- Frictionless File/Text Upload Wrapper: A responsive interface containing an input container titled "Paste your LinkedIn profile text or drop a Resume" to simulate top-of-funnel data gathering.
- 3-Step Interactive Preview Questionnaire: A lightweight form sequence capturing basic operational competencies (Sales, Operations, Development, etc.) to mirror the future AI onboarding experience.
- The Churn-Guard Paywall Overlay: A polished fallback modal triggered upon form submission to capture user intents without breaking the budget on backend AI servers.
- Database Endpoint Mapping: User inputs must securely map to the local database table schema titled
foundersradar_waitlist. - On-Device Masking Protocol: Before strings leave the browser, client-side regex scripts must scrub out obvious PII (Personally Identifiable Information) such as names, phone numbers, and physical addresses to maintain universal GDPR/CCPA data compliance.
- Intellectual Property Protection: Core promotional text and value strings must be rendered in native plain HTML/CSS arrays rather than structural JavaScript constants to block competitors from harvesting business rules via basic inspection utilities.
Because we utilize Kiro IDE, the development team must strictly adhere to the Spec-Driven Development flow. Do not write raw code without executing these sequential pipeline stages:
Create or edit the requirements file detailing explicit parameter bounds, input forms, and data sanitation parameters.
# Example Kiro requirement validation command
kiro spec:validate requirements.mdMap out component placement indices, state wrappers, and layout endpoints before generating files.
Break the development into individual atomic checkpoints. Run the development environment in parallel sandbox targets to isolate code side-effects.
- API Budget Hard Caps: Go to your individual developer portal accounts (OpenAI/Anthropic Dashboard) and hardcode a strict $50/month spending threshold limit on Day 1 to prevent accidental runaway code loop billing.
- Use Pre-Built Primitives Only: Do not try to program unique layout algorithms from scratch. Use Tailwind CSS classes and verified components to guarantee immediate security, responsiveness, and performance.
- Strict Data Decoupling: Keep business descriptions out of variable states. Hardcode typography strings directly inside structural elements to eliminate competitor scraping vectors.
| Column Name | Data Type | Constraint | Description |
|---|---|---|---|
id |
uuid |
PRIMARY KEY, DEFAULT gen_random_uuid() | Unique entry record key. |
created_at |
timestamp |
DEFAULT now() | Data creation entry timeline. |
email |
text |
UNIQUE, NOT NULL | Cleaned target user user email. |
profile_payload_masked |
text |
NULLABLE | Anonymized user professional text. |
competency_scores |
jsonb |
NULLABLE | Object recording core onboarding paths. |
The waitlist form submits masked profile text and competency answers to POST /api/waitlist. The route validates the payload server-side and upserts by email through Supabase using the service-role key. The service-role key must remain server-only and must never use a NEXT_PUBLIC_ prefix.
- Create a Supabase project.
- Run
supabase/migrations/001_create_foundersradar_waitlist.sqlin the Supabase SQL editor. - Copy
.env.exampleto.env.localand fill in the project URL and service-role key. - Start the app with
npm run dev.
Without those environment variables, the API intentionally returns 503 rather than reporting a false successful signup.
This project is prepared for both Vercel and Netlify.
- Push this repository to GitHub.
- Import the repo into Vercel.
- Set the environment variables from
.env.examplein the Vercel dashboard. - Use the default Next.js build settings.
- Add the repository to Netlify.
- Set the same environment variables in Site settings → Environment variables.
- Use the default build command:
npm run build. - The repo includes a
netlify.tomlfile so the Netlify Next.js plugin is configured automatically.
The local experiment currently includes:
- A deterministic matcher at
POST /api/matchesusing four curated startup patterns. - Ranked match results with fit scores, matched strengths, gaps, and six-month roadmaps.
- PDF, DOCX, and TXT resume parsing at
POST /api/parse-resume, with on-device/server-boundary masking. - A Supabase Auth-shaped sign-in screen at
/authfor the future magic-link integration. - Supabase migrations for profiles, saved matches, and match history.
- A local analytics event buffer and an LLM provider interface with a no-cost template fallback.
- Vitest coverage for the matcher and common PII masking behavior.
The sample startup patterns are intentionally illustrative. Replace them with reviewed, licensed data before presenting match scores as research-backed recommendations. The LLM adapter is also disabled by default; adding a provider requires a server-side key, privacy review, rate limits, and output evaluation.
Developed using Kiro IDE | © 2026 FoundersRadar. All Rights Reserved.