MLOps Predictions 🔮

When Adria Romero presented his project ProdutizeML at the MLOps community earlier this year, he started interacting with a lot of other community members and discussing various collaboration opportunities.

Among these members were Mateo Rojas-Carulla, CTO and founder of Lakera AI, a startup in Zurich working with how to develop, test, and operate safety-critical ML systems. Adria was one of Lakera’s first product testers, giving valuable feedback to the team–initially all through the MLOps community workspace.

After iterating through various product versions together, both Adria and Lakera decided to join forces: Adria became Lakera’s first software engineering hire: check out this post for proof!

Looking back, Adria has nothing but good things to say about how helpful the community was in making the right connections. He’s still a daily community visitor and contributor.

PS Lakera is still hiring, so maybe there is another success story in the making?

For those of us that don't live in the mighty country of hockey and maple syrup, let me introduce you to Loblaws. It's one of Canada's biggest grocery store chains. Being the ignorant US citizen that I am, I'd never heard of it.

Apparently, they are pretty big. Big enough to have a dedicated data platform team of 7 people that service the 60 data scientists working there. The use cases span the board from recommended systems to inventory prediction.

Here is a quick snapshot of the stack they are working with. Kafka, BQ, Airflow and Vertex AI. They also tried out the classic open-source tools like MLflow and Feast. Our guest Mefta explained why those tools didnt work out for them and so much more on this second espresso shot.