Blog
Honest notes from the Zafronix studio — building QORA, VenueFuze, ZKickoff, Siono, the Zafronix APIs, CrisisGraph, and Next-Gen Chess. What works, what doesn’t, what we got wrong. For posts about our games, see games.zafronix.com.
-
Why we built ZKickoff
Soccer competitions still lose time and trust to paper match reports and spreadsheets. ZKickoff makes the referee report digital — and standings, brackets, discipline, and public pages update from one verified record.
-
The Footballer, Referee & Club APIs
Three new football data APIs that go across competitions: every footballer’s career, every referee’s analytics, every club’s trophy cabinet — one canonical id each, in JSON, on one key.
-
QORA in the field: four jobs your cameras can already do
After-hours intrusion, vehicles at the gate, privacy-safe monitoring, and one operator console. Four real security jobs — and how QORA handles each on the cameras you already own.
-
Why we built CrisisGraph
The web has timelines and encyclopedias, but no single place that treats history as a causal graph. CrisisGraph maps what led to what — with clickable citations and an AI that walks the chain.
-
What have we learned from World Cup 2026 so far?
288 goals, 4 shootouts, a debutant nation unbeaten through the groups, and all three host nations gone by the Round of 16. The data, charted.
-
Edge AI for security cameras: when video stays on-premise and when it goes to the cloud
The right question isn’t “edge or cloud” — it’s which video, for which inference, for which decision, lives where. A practical decision matrix for security teams.
-
Stop asking if it's AI. Ask which part.
There’s no such thing as “AI”. There are four building blocks — LLMs, RAG, agents, MCP. Real products are mostly old software with carefully–chosen pieces swapped in at the right spots.
-
Inside the Zafronix APIs family
Four APIs. One auth. One shape. One bill. The data infrastructure under our consumer products — opened up to other builders. Starting with the World Cup family.
-
Why we built QORA
Most physical-security ops in 2026 are still humans watching grids of cameras. QORA puts AI in front of the monitor — capture-to-dispatch from minutes to seconds.
-
Why we built VenueFuze
A school gala still runs on five tools that don’t talk to each other. VenueFuze collapses that to one cart, one identity, one QR — and one consolidated, itemized tax receipt.
-
Five things we learned charting every World Cup since 1930
Brazil 1986 flew 22,000 km. Estadio Toluca sits at 2,666 m. Salenko scored 5 in one match. The data behind 23 World Cups, charted at the WC Explorer.
-
Why we built siono
Group chats are great at making plans. They’re terrible at making decisions. siono is the smallest possible tool that fixes that — and the discipline behind keeping it small.
-
Why we built a World Cup API
Every World Cup since 1930 as clean JSON. Sub-200ms responses, free tier with no card. Built for the indie devs, sports bloggers, and data scientists who’ve been stuck between scraping HTML and signing five-figure contracts.
-
92 years of World Cup data: what we found
~2,500 players, 23 tournaments, 6 confederations. Birth-month curves, post-WW2 squad-size growth, the surprising hemisphere split — with code samples you can run.
-
A 5-minute live-data lag, by design
Live sports data is a ladder — sub-second feeds cost five figures a month, and most projects don’t need that. Here’s how we deliver fast-enough live coverage for fantasy games, prediction trackers, and recap widgets.
-
Why we built Next-Gen Chess
Most chess apps optimize for rating. We built Next-Gen Chess to optimize for learning — a free AI coach that explains every move while you play, with a classroom layer for schools that need a safer way to teach chess online.
No posts in this filter yet.
More posts coming — notes from across the studio as we build.