How HackathonFinder finds hackathons and keeps listings honest

Hackathons are announced in a lot of places: listing platforms, college websites, club pages, social media posts and community groups. If you’ve ever tried to track them yourself, you know how easy it is to hear about a great one the day after registration closed. HackathonFinder exists to bring those announcements into one searchable place. Here’s how that works, and the rules we follow while doing it.

Where listings come from

We look across several kinds of sources:

• Hackathon platforms such as Unstop, Devfolio, Devpost, MLH and HackerEarth, where many events are listed.
• College and club websites, which often announce events that don’t appear anywhere else.
• Web search, to pick up events posted on smaller sites.
• Instagram and LinkedIn events, which reach us through submissions from organisers and users, partners, our browser extension and official APIs. We don’t scrape anyone’s accounts.
• You. Anyone can tell us about an event through Submit a hackathon.

We’re starting with Pune and online events and will widen coverage over time. We won’t catch every single hackathon, which is exactly why submissions matter.

The agents behind the scenes

A set of AI agents handles the repetitive work, each with one clear job:

1. Scout looks for new and updated events across our sources.
2. Extractor pulls out the details that matter: name, organiser, dates, deadline, mode, city, team size, eligibility, fees and prizes.
3. Deduper spots when the same event appears in several places and merges those into one listing.
4. Verifier checks the details against the official source and flags anything that doesn’t add up.
5. Tagger adds themes and labels so filters like AI/ML, fintech or free events work properly.
6. Writer writes a short, original summary of the event in plain language.
7. Publisher puts the finished listing live and updates it when details change.

Splitting the work this way makes each step easier to check and improve. A tagging mistake, for example, can be fixed without touching how events are found.

Why deduplication matters

One hackathon might be listed on a platform, announced on a college site and shared on social media, sometimes with slightly different names or dates. Without deduplication you’d see three cards for one event and wonder which is right. Deduper compares names, organisers, dates and links to group them, and the listing keeps a link to the official page so you always know where to register.

Verification and “last checked”

Event details change. Deadlines get extended, venues move and extra rounds get added. So every listing clearly shows two things:

• Its source, with a link back to where the information came from.
• A “last checked” time, so you know how fresh the details are.

We work hard to keep listings accurate, but the organiser’s official page is always the final word. We don’t host registration either: when you click register, you go straight to the organiser’s own page.

The data rules we follow

• We follow each site’s terms of use.
• We use official APIs and feeds wherever they exist.
• Instagram and LinkedIn events come from submissions, partners, our browser extension and official APIs, not from scraping accounts.
• We always link back to the original source.
• We write our own short summaries instead of copying organisers’ descriptions.

How you can help

If you organise a hackathon or spot one we’ve missed, add it through Submit a hackathon. If a listing looks wrong, tell us in the community so we can recheck it. A weekly digest and deadline reminders are coming soon; for now, save events to My hackathons to keep track of them. For more on the process, see How it works.