Databricks rolls Astra out to all 3,500 engineers: coding spend up 60%, complex tasks saturated by older models
pwendell · x · 2026-09-17
A Databricks engineer shares results of rolling the Astra coding assistant out to 3,500 engineers: it clearly beats their previous top models (Opus 5, Sol 5.6) on complex system-design and long-horizon tasks, but shows little gain on medium/low-complexity work, which older models have mostly saturated. Astra adoption raised overall coding spend 60% vs baseline. The rollout used a 200-user pilot with cohort experiments via their Unity Gateway, plus per-model sub-budgets that push engineers to reserve Astra for hard tasks while mixing cheaper models. No Astra-vs-Fable comparison yet due to data retention policies.
More from coding & agent
- Microsoft's MemSteer turns unlabeled user histories into agent memory supervision — MicrosoftResearch · 2026-11-02
- She fired herself from her own Instagram: AI agent runs the account, views up 82.3% — heyshrutimishra · 2026-11-02
- Sourcegraph: AI writes code faster than teams can review it, driving more outages — AI Engineer · 2026-11-02
- Tsinghua & Microsoft's HarnessMM cuts agent token use by up to 87.93% while raising success rate — rohanpaul_ai · 2026-11-02
- GitHub launches Agent HQ: Anthropic, OpenAI, Google coding agents in Copilot — emmanuelvivier · 2026-11-02
- DeepSeek open-sources Super-Agent, a coding agent that reads team chats to minimize human input — emmanuelvivier · 2026-11-02