Hapnin
Hapnin is an event discovery platform for Nairobi that brings events from different websites, ticketing platforms, posters, and other sources into one clean, searchable catalogue.
Events in Nairobi are spread across ticketing sites, social media posts, venue pages, posters, and individual event websites. Finding something to do often means checking several different places, and the information is inconsistent — titles change, dates are formatted differently, events get reposted, and the same event can appear on multiple platforms.
I started Hapnin to solve the data problem behind event discovery, not just build another events listing page.
The product combines a consumer-facing discovery experience with an ingestion system that collects event information from multiple sources, converts it into a consistent format, identifies duplicates, tracks where information came from, and gives me an admin interface for reviewing how that pipeline is performing.
The long-term goal is a platform where people can reliably discover what is happening around them while organizers eventually gain tools for publishing, promoting, and managing their own events.
How it works
Hapnin separates event discovery from the ingestion machinery behind it.
On the consumer side, users can browse, search, and open individual event pages from one consistent interface.
Behind the scenes, the ingestion pipeline can accept event information through:
event URLs supported event and ticketing websites event posters pasted text manually managed sources
Crawl4AI handles browser-based crawling where normal HTTP extraction is not enough. Gemini is used for structured extraction and normalization, while deterministic matching rules help prevent duplicate events from being created.
The admin pipeline provides visibility into sources, crawl runs, imported events, failed attempts, warnings, and other ingestion telemetry instead of treating scraping as a black box.
Engineering areas I worked on canonical event data modelling multi-source ingestion deterministic event deduplication Gemini-assisted structured extraction poster-to-event extraction Crawl4AI worker integration Supabase database, authentication, storage, and RLS ingestion telemetry and admin tooling consumer event discovery UI source freshness and attribution crawler security and request restrictions
The next stages expand the same event and venue data into richer organizer tools and ticket intelligence without rebuilding the underlying platform.
1 of 5
Related work
All workSoko POS
A powerful Desktop and online hosted SAAS Point of Sale system for powering any kind of business. Supports M-PESA, E-TIMS integration, POS terminals and handheld devices, Dashboards and Reports, and many more.
Circle
A two-sided platform connecting businesses that need digital growth and promotional support with people who can earn by completing campaigns and other opportunities.
Aegis-X
Privacy-first AI security platform that detects phishing and business email compromise without exposing your private data to external LLMs.