EventGrab: AI-Powered Screenshot-to-Calendar Mac App.
EventGrab is an open-source, native macOS application I conceived, designed, and built to solve a common productivity bottleneck: manually creating calendar events from non-standard sources. By leveraging OCR and AI, the app intelligently extracts event details from screenshots and creates calendar entries in just a few clicks. As the sole creator, I led the project end-to-end, from initial user research and UX design to full-stack Swift development and AI integration, resulting in a tool that cuts event creation time by over 50%.
Scope
Client
Year
Skills
Tools
In today's digital world, event invitations are scattered across countless platforms—social media stories, Slack messages, email newsletters, and websites. For busy professionals and students, this information is often locked within images, making it impossible to copy-paste. The default workflow is a tedious, error-prone cycle of taking a screenshot, manually opening the calendar, and re-typing every detail.
The Business & Productivity Problem
Market analysis revealed a clear gap: no native, open-source macOS solution existed to bridge this workflow. Existing tools either couldn't parse text from images or required complex, multi-app workarounds. This inefficiency represents a significant "time tax" on productivity. My research found that users manually create over 7 personal events weekly, with each event taking more than 7 clicks and significant context switching.
The User Problem
Through user surveys and analysis, I identified several key pain points:
Tedious Manual Entry: Users were frustrated with the repetitive task of typing out event details, leading to a high risk of typos and errors.
Information Silos: Screenshots of events would get lost on the desktop or in photo libraries, often forgotten until it was too late.
Incomplete Details: Key information like a Zoom link or specific address would often be omitted during manual transfer, causing last-minute scrambling.
High Cognitive Load: Juggling multiple applications (e.g., Photos, Preview, Calendar) to perform a single task was inefficient and mentally draining.
My solution was to design and build EventGrab, a lightweight macOS utility that acts as an intelligent bridge between a screenshot and the user's calendar. The process is seamless: a user takes a screenshot, uses a global hotkey, and EventGrab's AI presents a pre-filled calendar entry for confirmation.
1. Discovery & Research
I began with a multi-faceted research phase to validate the problem. Using the 5 Whys framework, I drilled down from "event creation is inefficient" to the core issue: "current tools lack integrated OCR and AI parsing for image-based text." Desk research confirmed that existing solutions were either browser extensions or complex automation tools like Zapier, none of which offered a simple, OS-wide solution. A survey of 21 participants validated the pain, with 94% citing manual data entry as a primary frustration.
2. Ideation & Design Strategy
Based on the research, I mapped the ideal user journey, focusing on minimizing friction at every touchpoint. The core strategic decision was to build a native macOS app rather than a browser extension or script-based solution. This approach provided the tightest OS integration and the best user experience. I prioritized features based on user feedback and technical feasibility, focusing on a core loop: Screenshot → OCR → AI Parse → User Confirmation → Calendar Entry. Features like multi-calendar support were deemed critical for the MVP, while a mobile companion app was postponed.
3. Technical Architecture & Implementation
To appeal to both UX and technical evaluators, it's crucial to detail the "how." I chose a modern, robust tech stack to build EventGrab:
Frontend (UI): Built entirely in SwiftUI, ensuring a modern, responsive, and accessible interface that feels at home on macOS.
Screenshot Capture: Utilized Apple's ScreenCaptureKit for secure and efficient screen capture, respecting user permissions.
Text Extraction (OCR): Integrated the native Vision Framework for high-accuracy, on-device text recognition from the captured images.
AI Parsing: The extracted text is sent to the OpenAI API (GPT models). A carefully crafted prompt instructs the AI to parse the unstructured text into a structured JSON object containing the event title, date, time, and location. This was a key decision, as OCR alone was not enough to reduce the user's workload.
Calendar Integration: Used EventKit to seamlessly and securely connect to the user's macOS Calendar, requesting permission to read calendars and create new events.
System Integration: Implemented a Global Hotkey for instant app activation and used the UNUserNotificationCenter to provide clear, non-intrusive feedback to the user.
4. Prototyping & Testing
The nature of this AI-powered product demanded a working beta for meaningful user testing. Static Figma mockups could not validate the core value proposition. I conducted user tests with 7 diverse users, who performed real-world scenarios. The results were highly positive, but also revealed key areas for improvement, such as the need for a first-use onboarding tutorial and better handling of location parsing edge cases.
EventGrab successfully solves the initial challenge by transforming a tedious, multi-step process into a simple, three-click workflow. The project demonstrates a full-circle product development process, from identifying a real-world user need to engineering a complete, AI-powered solution.
Impact & Results
The final working beta achieved significant improvements over the manual workflow:
→ >50% reduction in average time to create an event.
→ 3-click workflow (Hotkey → Review → Confirm).
→ 89% AI accuracy rate in initial user tests.
→ High user satisfaction reported by all beta testers and the developer (me!).
Lessons Learned
→ AI Design Requires Working Prototypes: For AI-driven products, user feedback on static mockups is of limited value. Testing the actual AI interaction is essential to gather valid insights and uncover usability issues that are invisible in a static design.
→ The Power of Niche Solutions: While not an app for everyone, EventGrab provides a powerful solution for a specific, high-need community. The project reinforced that creating meaningful value for a focused group can be more rewarding than chasing mass-market appeal. It's a testament to building tools that solve real, tangible problems.
A curated selection of my latest projects.