# Databricks — How our AI Research team uses AI Runtime

- Company: Databricks (databricks.com)
- Announced: 2026-10-15T21:37:44+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://www.databricks.com/blog/how-our-ai-research-team-uses-ai-runtime
- Record: https://forck.live/items/23138-how-our-ai-research-team-uses-ai-runtime
- Subject: Mosaic AI

Databricks AI Research uses the company's AI Runtime (AIR) to train and fine-tune models at scale, from small experiments to workloads spanning hundreds of GPUs. The team relies on built-in GPU resiliency, fast job launches, and integrations with MLflow and Unity Catalog to accelerate experimentation. Agent-driven research has increased training jobs 8x, with AIR supporting faster, automated experimentation.

## Evidence

Verbatim from https://www.databricks.com/blog/how-our-ai-research-team-uses-ai-runtime:

> Built-in GPU reliability, fast job launches, and integrations with MLflow and Unity Catalog help researchers experiment faster.

---

Record: https://forck.live/items/23138-how-our-ai-research-team-uses-ai-runtime
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md