Labmate

Get into a research lab in 14 days. Labmate finds the professors at your school who fits you, flags labs that just got funded, and makes sure you actually read their work before you email.

Comment

Inspiration

Research experience is close to required for med school and PhD applications, and a big edge for internships. But the only way in is cold-emailing professors, and most emails go unanswered because they're generic. Families pay research-mentorship programs $2,900 to $8,900 just to get a foot in the door. We're students who went through this, and we wanted to build the tool we wished we'd had.

What it does

  1. Lab Radar: pick your university and interests. Labmate reads three years of papers from your school (OpenAlex) and ranks every active researcher by fit, weighted toward the people who run labs.
  2. Money just landed: Labmate cross-checks NIH RePORTER and NSF awards. A lab that got a new grant weeks ago probably needs hands now, and almost no student knows to look for this.
  3. Their email, one tap: Labmate finds the professor's address in the author details of their own published papers (Europe PMC) and shows which paper lists it. Real public data, never guessed.
  4. Paper brief: any recent paper explained in plain English, with key terms and smart questions to ask.
  5. Understanding gate: before Labmate writes anything, you pass a 3-question quiz on the paper and write your own takeaway. The email is built around your words, addressed to the professor, and it never invents experience you don't have.
  6. Your inbox, not ours: open the draft in your own Gmail, Outlook or Mail app. Labmate never sends anything for you.
  7. Outreach tracker: follow-up reminders at day 7 and a cap of 5 emails a week, so professors get fewer, better emails.

How we built it

  • Expo (SDK 57) + Expo Router + TypeScript, for iOS, Android and the web.
  • Designed to Apple's Human Interface Guidelines for iOS 26: Liquid Glass (expo-glass-effect) on the controls only, SF Symbols, SF Pro and SF Pro Rounded, system colors with dark mode, native large-title headers and haptics.
  • Data: OpenAlex (researchers and papers), NIH RePORTER and NSF Awards (grants), Europe PMC (professor emails). The ranking is a pure, unit-tested function.
  • AI: structured outputs (zod schemas → JSON Schema) for briefs, quizzes and drafts, served from Vercel functions. It runs on Qwen through OpenRouter for about $0.0006 per brief (Claude is supported as a drop-in). Slow or broken answers are retried on another provider, and a brief is prefetched when the lab page opens so it's usually ready by the tap. Keys stay server-side; OpenAlex and NIH are proxied with caching, a parameter allowlist and a per-IP rate limit.
  • RevenueCat: entitlement pro, an offering with a 3-month Research Season Pass ($27.99) and a monthly plan ($9.99), a plan picker, restore purchases, and the customer-info listener so Pro unlocks instantly. Free tier: 3 briefs. The same RevenueCat setup runs in the web build.
  • Tooling: ESLint, Vitest (app + server tests), GitHub Actions CI on every PR, EAS build profiles.

How it uses RevenueCat

  • Seasonal pricing that matches how students actually search for labs: a Research Season Pass instead of a forever subscription.
  • A natural upsell: the fresh-grant details and unlimited briefs are gated behind pro, and free users can see that a lab just got funded.
  • Parent checkout: "Ask a Parent to Pay" shares a RevenueCat Web Purchase Link (RevenueCat Billing + Stripe) tied to the student's app user ID, so a parent can pay on the web and the student's Pro unlocks live.

Challenges we ran into

  • OpenAlex started requiring API keys in 2026, so we moved every call behind a caching proxy.
  • Matching the same person across OpenAlex, NIH, NSF and Europe PMC, which all format names and institutions differently.
  • Making an AI email tool that professors won't hate. The quiz gate and the weekly cap are product decisions, not just features.

Accomplishments that we're proud of

  • Works for any university on day one, with no partnerships and no scraping, using only open data.
  • The "money just landed" signal: public grant data turned into an unfair advantage for students.

What we learned

Monetization design is product design: who pays (parents), when (application season) and what feels fair (a season pass).

What's next for Labmate

  • Mock lab interviews generated from the lab's own papers.
  • A high-school edition: summer research programs and mentor matching.
  • Publishing to the App Store and Google Play before the spring research season.

Built With

Share this project:

Updates

Submission history