Halo

Inspiration

While building a B2B startup and attending conferences and meetups, I kept running into the same problem: I would leave with a stack of business cards, but a few days later, I could barely remember who each person was or why we should reconnect.

Existing business card scanners are good at digitizing contact details, but the relationship context still gets lost. I built Halo to turn every business card into a useful, searchable connection instead of another forgotten contact.

What it does

Halo is a digital business card and contact management app that helps professionals:

• Create and share a digital business card
• Exchange contact information through QR codes
• Scan paper business cards using AI-powered recognition
• Review and edit extracted contact details
• Organize contacts in a searchable archive
• Save the context behind each connection for future follow-up

The goal is simple: scan business cards, remember who you met, and follow up smarter.

How I built it

Halo is a mobile-first iOS app built with React Native, Expo, and TypeScript.

I used HeroUI Native to create a clean, modern interface and Supabase for authentication, contact storage, and backend services. The scanning workflow converts unstructured business card images into structured contact data while keeping the user in control through an editable review step.

I integrated RevenueCat to manage subscriptions and entitlements. This allowed me to build a free experience where users can try Halo's core value before upgrading to a Pro plan for higher usage limits and premium capabilities.

Expo Application Services helped me build, test, and ship the app to the Apple App Store.

Challenges I faced

Handling unpredictable business cards

Business cards have no standard layout. Names, job titles, phone numbers, logos, QR codes, and company information can appear anywhere. Cards may also mix Traditional Chinese and English.

The challenge was not only extracting text, but correctly mapping it into structured fields without silently saving incorrect information.

Designing trust around AI output

Even highly accurate recognition can make mistakes. Contact information is sensitive, and one incorrect digit can make a lead unreachable.

I designed a review step that lets users quickly verify and edit extracted information before saving it.

Building a fast capture flow

Networking interactions are short and sometimes awkward. The scanning experience needed to feel faster than manually entering a contact without interrupting the conversation.

I spent time reducing unnecessary steps between opening Halo, scanning a card, reviewing the result, and saving the contact.

Managing subscriptions reliably

Subscription state needs to remain consistent across purchases, restores, reinstalls, and account changes. RevenueCat helped centralize entitlement logic so the app could provide the correct experience without spreading purchase checks throughout the codebase.

Accomplishments that I am proud of

• Shipped my first iOS app to the Apple App Store
• Built the complete flow from scanning a physical card to saving a structured contact
• Combined digital business card sharing and paper card scanning in one product
• Added a real subscription system instead of postponing monetization
• Designed Halo as a usable product for real networking situations, not only a technical demo

What I learned

I learned that recognition accuracy is only one part of the product. A trustworthy review experience, clear error handling, and fast interaction design matter just as much as the underlying AI.

I also learned that retention for a networking product should not be measured by how often someone opens the app. The real value is whether users can remember, find, and reconnect with the right person when it matters.

Most importantly, I learned how much work exists between building a feature and shipping a complete consumer app—including subscriptions, privacy, onboarding, edge cases, and App Store distribution.

What's next for Halo

The current version focuses on digital cards, scanning, and contact organization. Next, I plan to explore:

• Event-based contact organization
• Follow-up reminders and AI-assisted outreach
• Conversation notes and relationship context
• Contact and CRM integrations
• Shared lead capture for teams
• Event and representative attribution
• An AI relationship assistant that helps users understand who they know and when to reconnect

My long-term vision is for Halo to become a personal relationship layer for the agent era—helping people preserve the context behind every real-world connection.

Built With

• eas
• expo.io
• heroui
• ios
• native
• ocr
• postgresql
• react
• revenuecat
• supabase
• typescript