Back to Portfolio
📇 Live Product · AI · Mobile · Automation

CardSync

AI-Powered Business Card Scanner

How StackLab eliminated 100% of post-event contact entry by building an AI system that scans, extracts, and automatically follows up with new contacts.

0s
Manual data entry per contact
Gemini 2.5 Flash
AI model powering OCR
Instant
Email + WhatsApp trigger

The Problem

Every sales professional, founder, and conference attendee knows the problem: a pile of business cards collected at an event that never gets properly entered into a CRM, and a string of introductory follow-up emails that never get sent because the manual data entry step is tedious enough to be perpetually postponed. Research suggests that over 80% of business cards are discarded within a week. The contacts stay on the table — literally — while business development opportunities evaporate.

The Solution

CardSync removes that friction entirely. A user photographs a business card with their phone camera. Google's Gemini 2.5 Flash model processes the image, extracts all relevant contact data — name, company, title, email, phone number — with high accuracy, and formats it for immediate use.

From there, CardSync triggers a personalized introductory email and WhatsApp message to the new contact automatically. Within minutes of a handshake, the contact is logged, a message is sent, and the user can move on to the next conversation. The entire post-event pipeline runs without a single manual step.

How It Works

01

Card Capture

User photographs a business card. The image is sent to the Gemini 2.5 Flash API with a structured extraction prompt.

02

AI Extraction

Gemini's multimodal capability processes the image, identifies and extracts all contact fields, and returns structured JSON — name, company, role, email, phone.

03

Contact Logging

The extracted data is saved to the user's contact database, with the card image attached for reference.

04

Automated Outreach

A personalized introductory email and WhatsApp message are generated and sent to the new contact automatically — without the user typing a single word.

Key Technical Decisions

Chose Gemini 2.5 Flash over GPT-4V for OCR tasks — stronger image understanding for card-format data at a fraction of the cost
Free-tier API architecture keeps inference costs near-zero for typical usage volumes
WhatsApp integration via Business API with templated introductory message formats
React Native for cross-platform iOS/Android deployment from a single codebase
Node.js backend with contact deduplication logic to prevent duplicate entries

Roadmap

Now
  • OCR contact extraction
  • Auto email + WhatsApp send
  • Contact database and card image storage
Next
  • AI-generated meeting notes per contact
  • Automated follow-up reminders
  • Team-shared contact database
Later
  • CRM integrations (HubSpot, Salesforce, Zoho)
  • Analytics on follow-up conversion rates
  • SaaS commercial launch

Tech Stack

Google Gemini 2.5 FlashReact NativeNode.jsPostgreSQLWhatsApp Business APIResend (Email)AWS S3
Chat on WhatsApp