AIdeas: GRÜM - Never get a bad haircut again.
Creating an "AI-enhanced Haircut Passport" for users, combined with an all-in-one business management tool for barbers anywhere in the world.
App category
"Daily Life Enhancement"
Team Members:
@grumbledore @rashinp
---
Alright, look, my idea isn’t curing disease or addressing multinational infrastructure issues (You guys are all awesome creators by the way, I’m so excited to see all of these ideas come to life). Alas, as the category says, I’m here to try and enhance your daily life by creating a solution that ensures You Never Get a Bad Haircut Again.
A bad haircut might seem like a small thing, but we’ve probably all had one, and we know that sometimes it can shatter your confidence for days or even weeks. This is even more true when you're traveling abroad or seeing a new barber and can’t quite explain what you want. GRÜM makes life easier by letting you save your favorite haircut so that you can get the same result anytime, anywhere. Just show what worked, share it with a tap or scan, and take the uncertainty and confusion out of the experience for both you and your barber.
The vision
The idea itself came from my own experience. When I was younger, I got haircuts very often, even weekly sometimes, and one bad cut could feel like a disaster, especially after spending my minimum-wage salary money and still leaving unhappy. Even as I got older and could afford better barbers, the same issue came up again and again: barbers are only human after all, and the good ones see hundreds of clients every week. Just walking in and saying “do the usual" doesn’t actually help either party at all and it’s an unrealistic demand.
GRÜM asks a simple question: What if you never had to get a bad haircut again?
GRÜM is an all-in-one mobile app for clients and barbers that makes great results easily repeatable and new styles easier to try.
GRÜM is an all-in-one mobile app for clients and barbers that makes great results easily repeatable and new styles easier to try.
Clients keep a private library of their best cuts, saved as a guided multi-angle photo set or short 360 capture, plus a standardized “cut recipe” and private barber notes. Next time they go for a cut, whether it’s the same barber or a new one, while traveling anywhere in the world, the client shares a time-limited QR code or simply taps their phone with the barber’s. The barber scans it and instantly opens a read-only Barber View with the exact angles, high-resolution 360 images, and instructions to replicate the cut as accurately as possible.
For barbers, GRÜM doubles as a portfolio and client memory system. Voice notes can be transcribed by AI and cleaned into consistent, reusable instructions that any barber can understand. With the client’s opt-in, finished cuts can be posted to a discovery feed so great work is searchable by style, hair type, and even location-specific trends.
Long-term, GRÜM expands beyond haircuts into a specialized beauty platform for consistent results, discovery, booking, payments, and certification across services like hair color, salon treatments, nails, lashes, and more.
GRÜM is a "haircut passport" that lets anyone get the same great cut anywhere, anytime.
Why this matters
The exciting part first
For the client:
- Never get a bad haircut again: save your best cut once, reuse it anywhere.
- Instant Barber View via QR: one scan opens angles + recipe, no awkward explaining.
- Guided multi-angle capture: a high-quality 360-reference that barbers can actually use.
- Standardized cut recipe: guards, fade/taper, top length, lineup, beard, products.
- “Social” style feed that shows other GRÜM users’ haircuts around the world, providing inspiration for clients and showcasing the barber’s best work.
- AI-powered haircut recommendations specific to customer face shape, age, and even geo-dependent beauty and fashion trends.
- Travel-proof: walk into a new shop in any country with the same clarity as your usual barber.
- Automatic translation of barber notes for cross-language appointments to reduce misunderstanding when traveling.
- Browse styles & barbers, try haircuts on your face with AI, schedule appointments, and pay in the same app.
For barbers and strategic partners:
- All-in-one platform where they can run their entire business (scheduling, payments, marketing)
- AI-generated detailed "cut recipes" from a simple multi-angle photoset or 360 video.
- Barbers can sell or “rent” their cut recipes to barbers anywhere in the world, giving them a passive way to earn income while simultaneously building their portfolio.
- Voice notes that clean themselves up: barbers speak naturally; the app formats it into clear instructions.
- Future R&D: AR overlays(smart glasses overlays), voice-guided cutting via headphones, full-session recording to auto-generate recipes, Bluetooth-linked smart clipper attachments & data transmission in real time.
- Future partner ecosystem: beauty brands can advertise to relevant style audiences in the discovery feed, opt-in, and privacy-safe.
And why people should care
GRÜM matters because miscommunication is the real cause of most bad haircuts. A blurry celebrity photo or “do the usual” is not enough. Trust me, I’ve tried it myself. Telling my barber to make me look like Cristiano Ronaldo (when he doesn’t have much to work with) doesn’t work the way it’s supposed to. This is especially true when hair texture, face shape, and regional barber techniques differ. A bad haircut can shatter your confidence at the worst time; before a wedding, a job interview, or any moment you just want to look your best.
This matters to barbers too by the way. Great barbers lose repeat business when expectations are unclear, and a client leaves disappointed. Even if the barber did their best with a weak reference, the final result is, sadly, but realistically, all that matters to the client. GRÜM gives barbers a clearer target, faster consultations, fewer reworks, and higher retention through reusable client history.
As for the “social” and “business-functional” aspect of GRÜM, unlike Instagram or Pinterest, this app is built specifically for this workflow. It combines inspiration and execution: reference, recipe, chair-side sharing, and on the barber's side: booking, payments, and verified reviews. GRÜM-certified barbers will also benefit from attracting more clients through the platform by giving potential browsing clients the confidence that they will achieve the result they are looking for.
Note: renderings are from a demo version of the app, use minor AI generation, and are not final
Some barbers may worry about “giving away” their best techniques. GRÜM can support permissioned recipe sharing where barbers choose what’s shareable, what’s private, and what’s only viewable by approved barbers. Long term, this can evolve into paid recipe unlocks or “authorized access” windows for traveling clients, while still protecting the barber’s relationship and brand.
Note: Absolute screenshot prevention is unrealistic on consumer devices, but access can be controlled via time limits, watermarking, and audit trails.
How we built this
GRÜM is being built as a mobile-first platform with a serverless, event-driven AWS backend so the chair-side experience stays fast even when media and AI workloads scale. The product is structured around repeatable haircut documentation, instant QR sharing, discovery, booking, and optional AI workflows that reduce barber effort.
Development approach
Mobile first: React Native 0.81 + Expo 54 (managed workflow) with TypeScript.
Serverless first: API Gateway + Lambda for most application logic, with ECS Fargate reserved for long-running media and batch AI jobs.
Event-driven pipeline: EventBridge and SQS decouple uploads, processing, AI analysis, and notifications so the app remains responsive.
Architecture overview
GRÜM has three core layers:
- Client and Barber mobile apps (React Native)
Separate navigation flows for client and barber roles.
Guided capture flow for front, left, right, back, and top angles.
Barber View is optimized for quick reference during the cut.
- API and real-time layer (AWS)
API Gateway HTTP API for REST endpoints (haircuts, bookings, feed, payments, AI).
API Gateway WebSocket for real-time events like booking updates, reminders, reviews received, and share link accessed.
Lambda function groups by domain (haircuts, media, bookings, feed, payments, AI, notifications).
- Data, media, and intelligence layer
Aurora PostgreSQL Serverless v2 as the primary relational database for bookings, payments, reviews, and feed posts.
DynamoDB for high-throughput workloads like WebSocket connections, engagement events, and real-time availability.
S3 + CloudFront for media storage and fast delivery, with signed URLs for private assets.
ElastiCache Redis for caching feed queries, shop lookups, and rate limiting.
Database design highlights
The schema is designed around durable “haircut entries” and structured recipes:
haircut_entries: user-owned haircut records with privacy controls.
cut_recipes: standardized recipe fields like fade type, guard progression, top length, lineup style, beard notes, and products used.
media_assets: angle-tagged photos and optional 360 video with CDN URLs and thumbnails.
share_links: time-limited, revocable QR tokens for chair-side sharing.
barbers, shops, bookings, payments, reviews: supports discovery, scheduling, and trusted reputation.
Media pipeline
GRÜM uses a direct-to-S3 upload pattern to keep uploads fast:
Mobile requests a presigned upload URL from the backend.
Mobile uploads directly to S3.
S3 triggers processing via EventBridge → SQS → ECS media-processor.
Processing strips EXIF metadata, generates thumbnails (200/400/800), optimizes assets, and updates DB records with CloudFront URLs.
Video uploads follow a similar flow with transcoding and thumbnail generation.
Video uploads follow a similar flow with transcoding and thumbnail generation.
AI integrations
AI is treated as infrastructure to reduce friction, not a gimmick:
Voice notes → recipe fields: barber voice recordings are transcribed with Amazon Transcribe using a custom barber vocabulary, then parsed into structured recipe JSON using Amazon Bedrock (Claude).
Photo analysis and moderation: Amazon Rekognition supports content moderation and can extract general visual labels that are then refined into haircut-specific attributes.
Style similarity search: Titan Embeddings generate style embeddings stored in pgvector, enabling users to find visually and structurally similar haircuts.
Recommendations: Bedrock produces personalized suggestions using non-identifying profile attributes and user history, with the results cached for performance.
Payments and payouts
Payments are implemented with Stripe:
PaymentIntents for client payments and tipping.
Stripe Connect for barber onboarding and payouts.
Tips can be forwarded fully to barbers while the platform takes an optional service fee.
Security and privacy
Cognito handles authentication with role-based access control for the client and the barber.
Haircuts are private by default, and public posting is opt-in per cut.
Private media is served via signed CloudFront URLs, and share links are time-limited and revocable.
EXIF stripping prevents accidental location and device metadata leaks.
Deployment and operations
Infrastructure is managed with AWS CDK across dev, staging, and production environments.
CI/CD runs through GitHub Actions for linting, type-checking, tests, and deployments.
Observability uses CloudWatch dashboards + alarms and X-Ray tracing across Lambda and API Gateway.
Cost note
This architecture is designed to scale efficiently. At ~10,000 MAU, the estimated infrastructure cost is roughly $470–$770 per month, with clear levers for cost optimization through caching, serverless scaling, and media lifecycle policies.
Demo
For a more in-depth look and a preview of the barber's view app (coming soon), check out https://rethink-grum.web.app/
What I learned
The hardest conceptual challenge is not storing photos; we’ve been able to do that for ages. It’s translating a haircut reference into guidance that reliably produces the same outcome across different barbers, hair textures, and tools. That’s why GRÜM prioritizes a structured recipe and consistent capture workflow before advanced AR features.
The biggest surprise is how feasible this is today compared to when I first explored the idea in 2017–2018. Modern mobile tooling and AI services drastically reduce the cost and complexity of prototyping guided capture, transcription, and structured extraction.
The key tradeoff is focus. The vision can expand into discovery, booking, payments, certification, and brand partnerships, but the MVP must prove the core functionality first: save what worked, share it instantly, and reduce miscommunication in the chair. If the workflow and data model are right, the platform can scale into bigger opportunities and become the global standard for this use case when the time is right.
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article