How it works
A hackathon discovery engine, not another registration form.
Organisers publish where they already publish. We watch those platforms, normalise what they say into one shared model, and give you a single place to search it — then hand you straight back to the original site to sign up.
How it works
A pipeline, not a spreadsheet
Hackathons are never entered by hand. The site reads from our own database; the database is kept current by a background sync that runs on a schedule.
- 01
Discover
Every six hours a scheduled job asks each connected platform for its current hackathon listings, preferring official feeds and respecting robots.txt and rate limits.
- 02
Normalise
Source-specific adapters map wildly different payloads onto one shared model: dates, prize pools, team sizes, eligibility and mode all become comparable.
- 03
Validate
Records are schema-checked before they reach the database. Anything malformed is dropped and reported instead of polluting the catalogue.
- 04
De-duplicate
The same hackathon on three platforms becomes one card. Registration URL, normalised title, organiser and dates are scored together to decide what merges.
- 05
Store
Results are upserted into PostgreSQL. Changed deadlines and prizes are recorded, and events that vanish are marked stale rather than deleted.
- 06
Redirect
We never take registrations. Every Register button sends you to the organiser's original page, exactly as they published it.
Questions
What the platform does, and deliberately does not, do.
Do I register through HackathonHub?
No. We are a discovery platform, not an organiser. Every Register button opens the original listing on the platform where the hackathon was published, with its registration URL preserved exactly as the organiser wrote it.
How often does the catalogue update?
A scheduled background job runs on the "0 */6 * * *" cron expression — every six hours by default — and can be triggered manually from the admin dashboard. Pages read from our own database, never from live scraping on request.
What happens when the same hackathon is on several platforms?
The deduplicator scores registration URL, normalised title, organiser and event dates together. Above the merge threshold the listings collapse into one card that shows every platform it appears on; borderline pairs are flagged for a human instead of being merged blindly.
What happens when a hackathon disappears from its source?
Nothing is deleted immediately. The event accumulates missed syncs and is marked stale once it has been absent from several consecutive runs, so a transient outage cannot wipe the catalogue.
How do you access the source platforms?
We prefer official or public machine-readable endpoints, identify ourselves with a descriptive user agent, honour robots.txt, apply per-host rate limiting and retries with backoff, and never bypass CAPTCHAs, authentication, paywalls or anti-bot systems. Where no compliant automated access exists, the adapter ships disabled and says so.
Is any of the data made up?
No. Every synced record keeps its source, source event ID and original URLs. The seed script used for local development marks its rows explicitly as demo data, and they are labelled as such in the interface.
Who built this
HackathonHub was built by Manan Maisheri, who kept running into the same problem as everyone else who enters hackathons: the building was never the hard part — finding the right event was. Every hackathon lives on a different platform, with its own listing page, its own filters and its own idea of what counts as a deadline.
Seven open tabs later, you would still miss something that closed while you were looking elsewhere. This site is the fix for that.