Skip to main content
Auto-memory
Retell + Twilio
Mortals
MCP ecosystem
Resume
Founder Pres.
Obsidian + QMD
Primary email
Review pass
Claude / Codex
Convex
75K+ sold
SwiftUI CloudKit
Leading Outdoor
Mission Control
AI Stack
ESR Jewel
Next.js + Vercel
Mortals
1,350+ locations
Shopify
Session notes
NightShift
Instagram
Founder
GitHub
MCP forks
<60s response
Toshi AI
Agent Infra
Full resume
NightShift
Operator chapters
Tally
QMD retrieval
DoubleTake
Roost
X
Founder & CEO
Luma
LinkedIn
cmux Inbox
121+ skills
Skill Sync
Six figures raised
Beam Mail
Whispa
Dragonfly Health
Memory System
Boswell
Founder office
DoubleTake
This portfolio
Full timeline
back to grid
Projects2026

PintSignal

An evidence-backed gluten-free beer directory proof of concept for web and iPhone that scans menus with camera OCR and looks up barcodes to distinguish dedicated gluten-free breweries from gluten-removed beers. Guidance is evidence tracking rather than medical advice, and trademark clearance remains pending.

Stack
POC
proof of concept
Vision OCR
on-device menu extraction
Fail-closed
unknowns never default safe
Web + iOS
Next.js and SwiftUI
§01The workLive system

An evidence-backed directory and mobile scanner checking gluten-free beer claims on web and iPhone.

The work

Check beer claims with transparent evidence.

PintSignal checks typed queries, menu photos, and barcodes against verified brewery evidence so drinkers know exactly what is in the glass.

  • Dedicated facilitiesSeparates dedicated gluten-free breweries from shared lines
  • On-device OCRExtracts printed menu text without uploading user photos
  • Fail-closedMarks missing or conflicting records unverified
Verification pipelineDedicated vs removedIngredient and facility evidence before any safety tag
01Scan menu
02Verify facility
03Flag risk
04Show evidence
Next.jsSwiftUIApple Vision OCRSupabase
The workPintSignal checks typed queries, menu photos, and barcodes against verified brewery evidence so drinkers know exactly what is in the glass.
§02The pipelineEnd-to-end flow
01

Scan menu

Apple Vision extracts beer names from taproom menus on the phone without uploading photos.

Apple Vision
02

Verify facility

Brewery records confirm whether production runs on dedicated gluten-free lines.

Brewery records
03

Flag risk

Enzyme-treated barley beers are clearly distinguished from naturally gluten-free grains.

Enzyme analysis
04

Show evidence

Drinkers see dated test results and facility notes before placing an order.

Evidence ledger
§03What I builtPieces that compound

Next.js web directory, SwiftUI scanner app, and deterministic classification rules.

On-device OCR

Menu text is recognized locally on the iPhone so personal photos never leave the device.

Apple Vision

Fail-closed defaults

Missing records and conflicting sources never default to safe.

Strict classification

Dedicated vs removed

Clearly separates beers brewed from gluten-free grains from enzyme-treated barley.

Facility tracking
§04Operating rulesEvidence bounds

Strict evidence bounds and fail-closed safety.

Verification rules4 of 4
  1. Provide factual evidence tracking rather than medical advice.
  2. Default to unverified whenever records are missing or ambiguous.
  3. Distinguish dedicated gluten-free facilities from gluten-removed beers.
  4. Process all menu camera images locally on the device.
§05StackEvery tool earns its line
Next.js / Web directorySwiftUI / iOS clientApple Vision OCR / On-device text extractionSupabase / Catalog databaseTypeScript / Shared schemas
§06ArtifactsReceipts and public proof

Web directory, iOS client, and domain schemas.

Web directory

Searchable catalog of brewery profiles and ingredient evidence.

Next.js web directory
SwiftUI iOS client

Native mobile app with on-device menu OCR and barcode lookup.

SwiftUI mobile app
Classification engine

Fail-closed rules that flag unverified or gluten-reduced beers.

Validation engine