Inspiration 💡

Applying to jobs can feel overwhelming--especially when you’re juggling multiple applications and trying to stand out in recruiters' inboxes. As early-career professionals, we noticed how tedious it was to personalize LinkedIn cold messages while also tailoring resumes for different roles. We wanted to make it easier and faster for people to connect meaningfully with recruiters by automating the process--without losing the human touch!

What it does 🔗

LinkedAI is a web app that helps job seekers generate personalized LinkedIn cold messages using their resumes. Users can upload their resume, which is parsed into structured JSON data. From there, the app suggests message templates tailored to specific roles and companies. Users can also save their message history and track whom they've reached out to--all in one place.

How we built it 👩🏻‍💻

We built LinkedAI using modern web development tools and AI APIs. The frontend is developed with React and Tailwind CSS for a clean, responsive interface. We used Supabase for user authentication and database management, enabling users to log in securely and store their message history. Furthermore, we also used Figma to design and prototype the user interface, helping us visualize the user journey and iterate on layout ideas before implementation.

For AI generation, we integrated Google’s Gemini API to parse resumes into structured JSON and generate personalized messages. Our Chrome extension interacts seamlessly with LinkedIn, allowing users to pull profile information and send tailored outreach messages with minimal effort!

Throughout the development process, we collaborated using GitHub, ensuring smooth version control and team coordination.

Challenges we ran into 😥

Parsing resumes was surprisingly tricky--formats vary wildly, and extracting accurate, structured data required lots of trial and error. We also had to balance the tone of AI-generated messages so they felt authentic and professional, rather than robotic. Integrating authentication and setting up a real-time database with Supabase also took time to debug.

Accomplishments that we're proud of 🙌

  • Successfully converting resumes into clean JSON data.
  • Building an end-to-end working demo with login, resume upload, and message generation.
  • Creating a design that’s simple and intuitive--even for users new to AI tools!
  • But most importantly, we are so proud of each member for giving their all and staying sane! Good job guys!

What we learned 🌱

We learned how to integrate multiple technologies across the stack--from frontend design to backend databases and AI APIs. We also developed a better understanding of user flows, especially for job-seeking tools, and how to keep UX friendly while handling complex processes like document parsing and text generation.

What's next for LinkedAI 🤔

We plan to:

  • Implement a feature where users can choose from or create message templates, so AI-generated messages follow a format they like or feel more personal.
  • Expand model support to include GPT and Claude, giving users more flexibility beyond Gemini and improving message variety and quality.

LinkedAI is just getting started-- we’re so excited to keep improving how people connect with opportunities! 😊🍌🐌

Built With

Share this project:

Updates