
How we increased or ML on Kubernetes capabilities and saved cloud costs
Improving ML capabilities on Kubernetes and reducing cloud costs

MLEnv: Standardizing ML at Pinterest Under One ML Engine to Accelerate Innovation
This blogpost discusses the implementation of MLEnv at Pinterest, a standardized ML engine that has resulted in improved ML developer velocity, ML platform productivity, and collaboration across teams.

What’s it like to write code at Meta?
Insights into writing code at Meta's scale and the experiences of software engineers at Meta.
Dear Duolingo: Are any words the same in all languages?
The blogpost explores two words that are universally shared by all languages and their significance.

Announcing Databricks Bengaluru Development Center
Announcement of the opening of Databricks' Bengaluru Development Center and the expansion of their R&D teams in India

Connection coalescing with ORIGIN Frames: fewer DNS queries, fewer connections
Exploring connection coalescing with ORIGIN Frames and its management at a large scale

Shared Clusters in Unity Catalog for the win: Introducing Cluster Libraries, Python UDFs, Scala, Machine Learning and more
Running more workloads on Databricks' multi-user clusters with security improvements.

What’s New for Shared Clusters in Unity Catalog
Enhancements to onboard more workloads to Unity Catalog clusters in shared access mode

Improving Worker Tail scalability
Enhancements to Workers Tail allow for enabling it at any size and scale

Elevating the generative AI experience: Introducing streaming support in Amazon SageMaker hosting
Introducing streaming support in Amazon SageMaker hosting for interactive generative AI applications, including chatbots, virtual assistants, and music generators.

Oxidizing OCaml: Data Race Freedom
OCaml with Jane Street extensions is available, with only a slice of the described features implemented.

Fast-tracking fusion energy’s arrival with AI and accessibility
Using AI and improving accessibility to accelerate the arrival of fusion energy with support from the DoE.