Loom | Devpost

expanded roadmap

Inspiration

We wanted to create something that could act like a personalized career guide for students, combining both local and online opportunities into a clear roadmap. Many students struggle to find internships, clubs, and learning paths tailored to their interests, especially in high school or early college. Loom helps bridge that gap by offering accurate, location-based recommendations and a progression system to keep users on track.

What it does

Loom takes a user's resume details, location, education level, and career interests to generate: • A personalized roadmap of opportunities throughout the year • Local internships, volunteer work, and research opportunities • College and major recommendations for high schoolers • Online certificates and bootcamps aligned with career interests • Nearby events like hackathons, maker spaces, or clubs based on zip code

It also provides a progression system so users can track growth and plan next steps, similar to having a private career counselor.

How we built it

• Frontend: We used Next.js to build a sleek, responsive web interface for user input and displaying results. • Backend: Built with Flask, handling: Experiences in parsing and profile analysis Integration with the Linkd API to search LinkedIn profiles and summarize experiences for recommendation insights • We also integrated Gemini for querying based on the user information provided to find the best opportunities for students alongside FetchAI, utilizing its agent-to-agent to decide how long to generate nodes for and when to stop

Challenges we ran into

• Many of us were using Next.js for the first time, so getting comfortable with its ecosystem was a hurdle • The Linkd API kept breaking during the early stages of development, but a big shoutout to Eric and the support team for helping us resolve those issues fast! • We had trouble getting the users' data through MongoDB, working through it, and eventually compromising with a global variable for Flask to use

Accomplishments that we're proud of

• We successfully created a project that links user data with meaningful, actionable opportunities • Integrated local and online resources into a flow-based roadmap • Developed a progression system that adds a gamified, trackable experience • Built a solid front-to-back stack using technologies we were mostly unfamiliar with at the start

What we learned

• How to work with Next.js and build dynamic, server-side rendered React apps • Better API handling and integration, especially with external data like LinkedIn profiles • Improved our teamwork under pressure, learning how to divide tasks efficiently between frontend, backend, and research • The value of personalized guidance in career development tools

What's next for Loom

• Enhance the recommendation engine with better AI/ML models or refined prompt engineering for even more personalized insights • Integrate more APIs for richer local data (e.g., Meetup, Eventbrite for clubs/events) • Essay critique, larger alumni network dataset • Create a custom message and create potential matches for networking