Browse past weeks of engineering reads.
Cloudflare needed to enable developers to automatically undo or compensate for failed steps in multi-step workflows executed across distributed systems.
Cloudflare needed to make OAuth available to all developers in their app ecosystem while migrating their core OAuth engine to a new implementation without causing downtime.
Cloudflare needed to enable multiple agent frameworks to build on top of their infrastructure without being locked into a single framework implementation.
Organizations need a way to programmatically plan, deploy, and manage Zero Trust security environments without requiring manual migration efforts or specialized domain expertise.
AI agents cannot programmatically deploy and run code on serverless platforms because authentication and account management systems were designed for human users, creating friction when agents need instant, automated deployment capabilities.
How to enable autonomous agents to programmatically create Cloudflare accounts, purchase domains, and deploy infrastructure without manual dashboard interaction or credential handling.
Enable multi-tenant platforms to execute millions of unique, durable workflows without incurring significant idle infrastructure costs.
Enabling AI agents to send, receive, and process email natively as a multi-channel communication medium without requiring developers to build custom email infrastructure.
Developers needed a unified way to access multiple AI model providers without managing separate integrations and API contracts for each one.
Cloudflare needed to enable enterprise customers to manage multiple accounts and resources under a unified organizational structure with centralized authorization and access control.
WordPress plugins pose significant security risks because they run with unrestricted access to the entire system, requiring a safer plugin architecture that isolates untrusted code.
Organizations struggle to migrate from legacy network security architectures to modern SASE (Secure Access Service Edge) solutions, facing risks from accumulated technical debt and complex dependencies in their existing infrastructure.
Running large AI models for agent workloads on edge infrastructure was cost-prohibitive and required significant inference stack optimization to serve models like Kimi K2.5 efficiently at scale.