Engineering
-
ServiceNow Published 64 Pages on People Intelligence. One Line Buried Inside Changes Everything.
ServiceNow just published a 64-page playbook on people intelligence. One buried statistic – only 24% of organisations have formal data cleaning processes – says more than the rest of the…
-
Why your agent works on Tuesday: notes from a team building LLM features into a multi-tenant SaaS
We are a small team shipping LLM features into a multi-tenant SaaS. The model is rarely the part that breaks. Capacity, schema drift, prompt caching, tenant isolation – the work…
-
One model won’t fit all: what we learned about churn prediction in People Analytics
We built a universal churn model for all our clients. It had a reasonable AUC. Then we ran a backtest and it caught zero actual departures. Here’s what happened next…
-
Ten years of multi-tenant SaaS: the decisions we made, and the ones we’d unmake
GFoundry. What held for ten years, what didn’t, and what a small team learns when decisions made in year two still cost you in year eight.
-
Stream mining in real time: what we built with Fraunhofer and what we learned
A €378,000 R&D project with Fraunhofer Portugal. MongoDB, ActiveMQ, FP-Growth, K-Means. What we built, what the data showed, and how it became GFoundry.
-
How we built a music recommendation engine in 2007 – before Spotify existed in Portugal
Palco Principal. 70,000 tracks, three engineers, no library worth using. The algorithm, the A/B test numbers, and how the 2008 financial crisis ended it.
-
Three platforms, one team: shipping for Heathrow Airport
iPhone, BlackBerry, Nokia QT — simultaneously. January 2012. What cross-platform mobile development actually cost before any of the frameworks existed.
-
What breaks when 20,000 artists depend on your ranking algorithm
Palco Principal had 20,000 artists, 350,000 monthly visitors, and a ranking algorithm none of them could see. Then the 2008 crisis hit. A failure story.