Careely

Careerly — Your AI Career Opportunity Companion Inspiration Students often miss valuable internships, hackathons, jobs, competitions, and programmes because opportunities are scattered across many platforms and deadlines are difficult to track. We wanted to create one place where a student can discover opportunities that actually match their interests, save them, track their application progress, and return to the original source to apply. Careerly was inspired by the idea of turning opportunity hunting from a repetitive manual task into a guided, personalized career journey. What it does Careerly is an AI-powered career dashboard that helps users: Discover live internships, jobs, hackathons, and competitions Search opportunities by role, skill, company, or domain Filter listings by source, including Internshala, Unstop, Remotive, Arbeitnow, and Greenhouse View the original source and apply through direct links Upload a resume for skill extraction and personalized recommendations Save opportunities and track their status Monitor deadlines, application stages, and next actions Receive a foundation for daily refreshes and future email notifications How we built it We built Careerly as a full-stack web platform using: Next.js and React for the user interface TypeScript for reliable frontend and backend logic Supabase for profiles, opportunities, applications, tracking records, and future notification data Vercel for deployment Public opportunity connectors for live listings OpenAI API support for resume parsing and skill extraction The platform uses connector logic to normalize opportunities from different sources into one common format. Each result includes a title, organization, type, skills, location, source name, deadline information where available, and a direct source URL. We also designed the system to keep source data fresh, avoid showing static demo opportunities as real results, and filter out listings that are explicitly marked closed. Challenges we faced The biggest challenge was working with different opportunity platforms. Every source has a different page structure, data format, update frequency, and policy around automated access. For example: Some platforms provide public APIs, while others expose only public listing pages. Listing HTML structures can change, which can break a connector. Some listings do not provide a deadline, location, or skill tags. A strict recommendation score could hide real opportunities, so we changed matching to rank results instead of removing valid listings. We had to ensure each card showed a visible source label and a working “View & Apply” action. Another challenge was balancing freshness with performance. Loading all sources can take longer than loading a single API, so we added controlled refresh logic and source-aware filtering. What we learned Through Careerly, we learned: How to build a scalable data model for opportunities, applications, deadlines, and user progress How to integrate multiple external sources into a normalized format How to create practical recommendation logic from skills, goals, and profile information Why source transparency matters when presenting live information How to handle incomplete, changing, and inconsistent real-world data How to deploy and maintain a production-ready Next.js application What’s next We plan to expand Careerly with: More approved opportunity sources and official APIs Gmail deadline and application-status notifications Better resume-to-role matching Application form assistance A calendar view for deadlines Personalized weekly opportunity digests Program tracking for opportunities that are not open yet

Built With

• actions
• ai
• api
• automation
• css
• github
• html
• javascript
• next.js
• node.js
• openai
• parsing
• postgresql
• react
• rest
• resume
• scraping
• supabase
• tailwind
• typescript
• vercel
• web