Latest

Every published item, in order · 16548 items
Nov 6
ProductsBuildersSelected01:33
Replit rebuilds sandbox engine around microVMs for AI agents

Isolating agents in per-sandbox microVMs with network egress controls and credential brokering addresses the security gap that stops teams from giving agents real production access. Platform and infrastructure engineers building agent runtimes can borrow the pattern of never handing agents durable credentials. It positions Replit's agent hosting against competitors relying on shared container sandboxes.

Official accountReplit@ReplitSource
Tomorrow 08:00, the daily lands in your inboxTelegram channel or email. Same screening as this page, no ads.
ProductsBuildersSelected01:17
Apache Spark 4.3 ships in Databricks Runtime 19

The quarterly release cadence is the quiet headline: Spark teams can expect completed features to land sooner instead of waiting on long release cycles. Data engineers running real-time pipelines should evaluate the new streaming aggregation and transformWithState APIs against existing Flink or Kafka Streams setups. Availability comes via Databricks Runtime 19, so upgrades happen through the platform rather than a standalone download.

Event
Official accountDatabricks@databricksSource
ProductsBuildersSelected01:05
Nous Research launches Hermes Agent Plugin Hackathon with $20,000 prizes

Judges explicitly favor plugins that go beyond what a prompt or skill could do, so winning entries will likely operate tools, memory or hardware rather than wrap existing abilities. Builders targeting agent marketplaces can study the scoring rubric — usefulness, reliability, polish and creativity — and the open plugin catalog doubles as a library of working examples once winners ship.

Official accountNous Research@NousResearchSource
ProductsBuildersSelected00:47
Together AI Launches Together Compiler for Polymorphic Inference

Automated compilation that replaces weeks of per-chip hand tuning could lower inference costs for anyone serving open models without a dedicated kernel team. Teams running self-hosted inference should evaluate whether the overnight optimization workflow fits their stack, especially against vLLM- and SGLang-based serving setups. Pricing and regional availability were not disclosed in the announcement.

Official accountTogether AI@togethercomputeSource
ProductsBuildersSelected00:37
Mistral launches Le Chonk model and Studio managed agents

The pairing of a new model with a managed agent platform signals Mistral's push to compete with OpenAI and Anthropic on enterprise agent infrastructure rather than models alone. Teams evaluating agent platforms gain another option with self-directed deployment, relevant where data residency or EU hosting matters. European availability and deployment control distinguish it from US-hosted alternatives.

Official accountMistral AI@MistralAISource

Only the latest 2000 items are listed; for older ones see the daily archive