Inspiration
Students don't always miss opportunities because they lack talent. Sometimes, they miss them because the information is difficult to understand or act on.
Internships, scholarships, hackathons, competitions, placement notices, and university opportunities are often shared as long announcements filled with eligibility rules, deadlines, requirements, and formal language. Students may have to read through everything just to answer a few simple questions:
- Am I eligible?
- When is the deadline?
- What do I need to submit?
- What should I do next?
- Can someone explain this in a language I understand better?
This problem becomes even more significant for students facing language barriers or who are unfamiliar with formal opportunity announcements.
That inspired us to build CampusBridge AI — an AI-powered platform designed to turn confusing opportunity information into clear student action.
Our idea is simple:
Talent is everywhere. Access to opportunity isn't. CampusBridge aims to help close that gap.
What it does
CampusBridge AI transforms an unstructured opportunity announcement into a clear, structured, and actionable student-friendly experience.
A student can paste an internship, scholarship, hackathon, competition, placement, workshop, or university notice into CampusBridge.
The platform then uses AI to extract and organize important information such as:
- Opportunity type and organization
- Eligibility requirements
- Deadline and important dates
- Mode and location
- Cost, prizes, or benefits
- Required documents and submission requirements
- Skills or other requirements
- Missing or unclear information
- Recommended next actions
But CampusBridge goes beyond summarization.
Students can also check their personal eligibility by providing basic profile information. CampusBridge compares their profile against the requirements explicitly stated in the announcement and explains each requirement as a match, no match, or unknown.
Students can ask questions about the opportunity, such as "What do I need to submit?" or "Can I participate?" Answers are grounded in the original announcement. If the information isn't available, CampusBridge says so instead of inventing an answer.
The platform also provides student-friendly explanations in Telugu, Hindi, Gujarati, Tamil, and Simple English, helping reduce language and comprehension barriers.
CampusBridge is designed around one principle:
Don't just summarize the opportunity — help the student understand what to do next.
How we built it
CampusBridge AI was built using a lightweight architecture so we could focus on reliability, accessibility, and the user experience.
We used:
- Python for the application logic
- Streamlit for the interactive web interface
- Featherless.ai for AI inference
- DeepSeek-V3.2 as the language model
- SQLite for saved opportunities
- Git and GitHub for version control and source-code management
- Streamlit Community Cloud for deployment
When a user submits an announcement, CampusBridge sends carefully structured instructions to the language model and requests a predefined JSON representation of the opportunity.
That structured result powers the rest of the application, including the visual opportunity cards, eligibility reasoning, multilingual explanations, next actions, and grounded Q&A.
We also designed the prompts to explicitly distinguish between information stated in the announcement and information that is missing.
Instead of allowing the AI to guess, missing information is marked as "Not specified" or identified as uncertain.
This makes the AI not only useful, but more transparent and trustworthy for students.
Challenges we ran into
One of our biggest challenges was getting reliable structured output from the language model. LLMs don't always return perfectly formatted JSON — sometimes responses contain Markdown code fences or unexpected formatting. We created a safer parsing layer that cleans the response, extracts the JSON object, validates expected fields, and handles malformed responses gracefully instead
Log in or sign up for Devpost to join the conversation.