A collection of the things I have built.

Chécalo keeps about 1.8 million products up to date from live chain data. Scraping them is the easy part. The hard part is deciding that a 1.5L bottle written three different ways in three different catalogs is the same bottle.

The backend is a distributed scraping pipeline feeding a matching layer that reconciles products across chains. Search runs on Meilisearch, Cloudflare handles the edge, and the app itself is React Native.

It works end to end. Ads and subscriptions are designed but not shipped yet. It is a commercial product, so the code stays private.

~1.8M

products indexed

Live

price updates

• Python
• FastAPI
• PostgreSQL
• Meilisearch
• React Native
• Cloudflare

• Visit checalo.mx

It started small. One police vehicle streaming telemetry at 10 Hz, an LLM reading tactical scenarios in plain language, and real routing over Madrid. That version won the national stage.

For the global final we grew it into a five fleet platform. Telemetry, island events and weather all move over Kafka. The dispatch dashboard updates live with OSRM routing and ETAs that shift as conditions do, and you can replay any past incident window.

Built with 52Sec. Third in the world.

10 Hz

telemetry rate

5

fleets simulated

• Python
• Flask
• Apache Kafka
• Docker
• MapLibre
• OSRM

• Visit HPE_Final
• Visit gemelodigitalhpe_52sec