Version 0.79.2 adds Google Vertex AI support to Activepieces. Builders should prepare for the mandatory secret manager refactoring coming in 0.80.0.

Hygraph's new MCP integration lets you update localized content through Claude conversations. Skip the dashboard - manage content via AI.

Cognition AI expanded Devin to manage multiple instances hierarchically. Here's what this means for your architecture and when you should adopt it.

DigitalOcean's new prompt caching feature cuts latency and inference costs by reusing cached context across API calls. Here's how to leverage it.

DigitalOcean now offers AMD Instinct MI350X GPUs, giving builders a viable alternative for AI workloads. This shifts the economics of GPU-dependent projects.

Apify releases MCP server enabling AI agents to extract data from thousands of websites via pre-built Actors. This standardizes web scraping for AI workflows.

Zapier now lets builders create custom actions for existing apps using AI assistance, expanding integration capabilities without requiring new app connections.

Vercel's new Sandbox capability enables safe execution of user-submitted code at scale. Notion is the first major platform adopting it - here's what builders need to know.

Zapier now lets you route AI inference through your own cloud provider accounts. Here's what builders need to know about cost control and infrastructure flexibility.

Vercel eliminates embeddings from knowledge agent architecture, reducing latency and infrastructure costs. Builders can now simplify RAG systems with direct LLM approaches.