gekro
GitHub LinkedIn
News

AI News

Salesforce-NVIDIA reasoning model targets enterprise tasks; Apple ships on-device Siri with Gemini

Salesforce released Koa, an open-weight reasoning model built on NVIDIA's Nemotron for enterprise workflows, while Apple deployed its rebuilt Siri with hybrid on-device and cloud inference using.

1 min read 5 sources

Salesforce and NVIDIA released Koa, an open-weight reasoning model built on NVIDIA’s Nemotron architecture and trained for sales, marketing, and customer-support tasks (TechCrunch). The model represents a vendor-specific deployment of open-weight infrastructure for enterprise automation, targeting domain-specific reasoning rather than general-purpose capabilities.

Apple shipped a rebuilt Siri with iOS 27, powered by Google’s Gemini models and running partly on-device through Private Cloud Compute, with early testers reporting improved multi-step request handling and screen context awareness, though some hallucinations and gaps in personal context remain (The Decoder). The EU does not receive the update due to regulatory constraints. In separate infrastructure work, Meta introduced ZGateway, a stateless proxy tier for its ZippyDB key-value store that handles over 1 billion operations per second (MarkTechPost). Agent-net open-sourced Webagent, a Go harness for converting websites into AI agents capable of inter-agent communication (MarkTechPost), and MarkTechPost published a practitioner’s reference on agent stack layers - harness vs. framework vs. MCP - and their ownership of control flow, state, tools, and permissions (MarkTechPost).

Compiled automatically from the linked sources and published without manual editing - a neutral summary of third-party reporting, for information only. Every claim links to its origin. Not original reporting.