Our brands · Reel In
Stop scrolling. Start watching.
One confident pick for tonight, built from your taste, not the algorithm's.
Visit reel-in.comarrow_forward
Most streaming apps solve the wrong problem. They surface infinite options, then call that a feature. Reel In asks a few quick questions and commits to a single recommendation. For people who can't decide, that's the whole point.
Two pillars
How it learns what you like
Letterboxd import
Upload your Letterboxd rating history (CSV or ZIP) and a taste profile gets seeded instantly. No blank-slate cold start.
Head-to-Head (H2H)
A quick pick-the-winner matchup between two movies. New users play a 25-round onboarding game. Everyone gets a fresh 10-round game daily, with a shareable result card.
Core features
Everything built around one decision
- theaters
Watch
The central recommendation flow. Answer a quick mood check and get one confident pick: solo, or a group pick using a least misery approach so nobody in the room hates the choice.
- video_library
Catalog
Browse popular, trending, and classic titles, or search the full movie database.
- bookmark
Lists
A curated watchlist capped at 5 titles per genre, forcing real curation over endless hoarding. Plus an Already Seen log that merges Letterboxd history with titles marked manually.
- group
Profile & sharing
Opt-in group taste sharing, so friends, roommates, or couples can get a blended recommendation together.
The recommendation engine
How the pick gets made
| Input | What it captures | How it's weighted |
|---|---|---|
| Taste affinity | Per-genre affinity from star ratings and H2H picks | About 65% of the blend. |
| Quality score | Critic and audience scores | About 35% of the blend. |
- Solo sessions get one confident pick, not a list to scroll through.
- Group sessions use a least misery strategy: optimized for the option the least-enthusiastic person still likes, not an average of everyone's taste.
- New users start with a 25-round H2H onboarding game. Everyone gets a fresh 10-round game daily.
- A 5-title-per-genre cap on watchlists keeps curation honest.
Where the data comes from
Built on real catalog and critic data
The catalog comes from TMDB. Critic and audience scores come from IMDb and Rotten Tomatoes. And anyone with a Letterboxd account can import years of rating history in one upload, seeding a taste profile that would otherwise take weeks of H2H rounds to build.
What makes it distinctive
Five angles on the same idea
- tune
Taste-first, not engagement-optimized
Built to make a decision, not maximize time in the app.
- check_circle
One pick, not a wall of options
A single recommendation per session, never an infinite scroll dressed up as a feature.
- groups
Group-aware by design
Built for couples, roommates, and friend groups deciding together, not just solo browsing.
- filter_alt
Curation enforced by design
A 5-per-genre watchlist cap forces real curation over endless hoarding.
- sports_esports
Onboarding that's actually fun
The first thing a new user does is play a game, not fill out a form.
Tech Stack
Built for real users, not a demo
| Framework | Next.js 16 (App Router), React 19, TypeScript, Tailwind CSS 4 |
| Auth | Firebase Authentication (email/password + Google sign-in) |
| Recommendation engine | Python FastAPI microservice, plus a Rotten Tomatoes score scraper |
| Database | PostgreSQL on Cloud SQL, via Drizzle ORM |
| Infrastructure | Google Cloud Platform: Cloud Run, Cloud SQL, Cloud Scheduler, Secret Manager, provisioned via Terraform |
| CI/CD | GitHub Actions |
| Data sources | TMDB, IMDb, Rotten Tomatoes, and user-imported Letterboxd history |
Status
Live, and built for what's next
Reel In is live at reel-in.com. Nightly cron jobs keep ratings fresh across the catalog, staging runs fully separate from production with its own database and pipeline, and the build has shipped in 14-plus structured workstreams: auth, catalog, recommendations, the H2H game, lists, onboarding, profile, and the production deploy itself.
Our brands
Explore what else we've built
USA CPG
Independent research and honest product rankings for the U.S. consumer-packaged-goods market. Every score, explained.
Read morearrow_forwardThe Local Archives
A field guide to the old buildings still standing in a city, written by us, kept honest by the locals who visit them.
Read morearrow_forward
Want a product like this for your business?
Recommendation engines, onboarding that doesn't feel like a form, real infrastructure under it. We build this kind of system for you, too. Let's talk it through.
Let's Talk