Tag
Databricks Vice President Nikita Shamgunov believes Claude Code is just the beginning. The AI software factories of the future will require entirely new agent infrastructure, and state management (such as database state) will be one of the hardest problems to solve.
MiniLake 是一个开源的本地 Databricks API 模拟器，提供真实的 SQL 执行（DuckDB）、真实的 Delta Lake 文件和真实运行的 Spark 作业，可在无云账号、无 API key 的情况下驱动 databricks-sdk、CLI、Terraform、Asset Bundles 及 MCP 智能体。
Databricks introduces MemCurrent, a persistent memory architecture for production agents that powers Agent Bricks Memory and Genie, combining plugin, tool, and offline "dreaming" pipelines to triple pass rates while cutting tokens by 31% and latency by 26%.
Databricks has moved Genie App Builder into Beta, allowing users to build governed, data-aware apps from natural language prompts inside the Databricks UI. Apps run on a new serverless micro apps runtime with scale-to-zero, are governed via App Spaces and Unity Catalog, and are free during Beta on AWS and Azure.
awesome-jev 的定期巡检新增 3 条条目（555 → 558）：llama.cpp 新增 System One 决策模型运行时支持、Databricks 推出 ai_decide AI Function 作为结构化决策 API，以及一个基于 Jev 官方文档的实际 RAG 测试案例。
The Genie One Model Context Protocol (MCP) server from Databricks is now generally available, enabling agents to access and query company data with built-in governance and support for various AI clients.
This article introduces how to use SQL on the Databricks platform to run the open-source decision model SemIf-OpenJev, enabling batch processing by calling the model without data leaving the platform.
Databricks announces sponsorship of Cal Bears Field, with all co-founders being UC Berkeley alumni, highlighting the company's roots in data and AI at Berkeley.
A tweet praises Ali Ghodsi, CEO of Databricks, for his technical and charismatic leadership qualities in Silicon Valley.
Databricks has opened up its Astra platform to approximately 3,500 engineers and drawn valuable conclusions from the initiative.
Databricks has rolled out GPT-6 Astra to all its engineers, claiming it outperforms Claude Opus 5 on complex tasks and boosts coding efficiency.
Databricks rolled out Astra, an AI model that outperforms previous models on complex tasks and increases coding spend, but shows no significant improvement on medium/low complexity tasks. The deployment includes budget controls to encourage selective use for cost efficiency.
A tweet discusses how Databricks has simplified Spark, enabling non-specialists like GTM Engineers to build data pipelines without deep technical knowledge.
GLM-5.3 achieves 310 tokens per second on Databricks inference, leading in both speed and latency, and is the strongest open-source model for coding, competitive with Fable 5 and Opus 4.8.
@swyx discusses Databricks' fundraising meme, linking their growth to AI agents starting to work in the enterprise, as highlighted by CEO Ali Ghodsi.
Databricks raised $5 billion at a $190 billion valuation after investor demand hit $15 billion, despite originally seeking only $1 billion. CEO Ali Ghodsi cited $7 billion annualized revenue growing 80% and heavy AI investment costs as reasons for the raise.
Elon Musk announces that Grok 4.6 reached #1 on Databricks, with Ivan Zhou reporting SOTA performance on OfficeQA Pro V2 using Databricks's Genie harness.
Databricks shares proven techniques for managing AI coding costs at scale, including moving to more efficient open-source models and using AI gateways, citing a 70% reduction in spend. The post covers strategies from Databricks, Stripe, Coinbase, Uber, and Ramp.
Databricks' benchmark shows that the same model invoked through different harnesses has a cost difference of more than 2x, while Pi, as a minimalist coding harness, delivers high performance at low cost; Shopify also used Pi to extend Autoresearch and improve efficiency.
Earendil's Pi coding harness demonstrates that minimalist design outperforms complex alternatives in cost and performance, citing Databricks benchmarks and a Shopify case study as evidence.