Diagrid, developer of the world’s most reliable foundation for mission-critical workflows and autonomous AI agents and the company behind the Distributed Application Runtime (Dapr), today announced Diagrid Catalyst 2.0, bringing durable and verifiable execution to the AI frameworks developers already use, including LangGraph, Microsoft Agent Framework, Google Agent Development Kit (ADK), AWS Strands, OpenAI Agents SDK, CrewAI and more. Teams can now make their existing agent workflows resilient to failure and cryptographically provable, without re-architecting how those agents are built.
AI agent failures are expensive. Most agent frameworks lack a built-in way to guarantee that a long-running workflow recovers and completes the job when something breaks, which means humans step in, tokens get burned, and projects stall. Diagrid Catalyst 2.0 solves this by combining durable execution with verifiable execution: agents automatically resume from the exact point of failure, and every step of that execution can be cryptographically signed and traced back to its source.
Catalyst 2.0 is a multi-cloud and sovereign solution for agent durability and cryptographic attestation that can run in fully air-gapped environments while improving the performance of open-source Dapr by up to 10x, enabling organizations to run millions of concurrent agent workflows. Diagrid is the only company that combines durable, verifiable execution into a single workflow engine built to take agents into production anywhere.
Developers add a Diagrid code package directly into their framework of choice and gain durable workflows immediately. LangGraph, for example, offers strong orchestration but only basic checkpointing for recovery, leaving the hard problem of detecting and recovering from failures to developers; paired with Catalyst 2.0, those same workflows now run to completion even after a failure, with no custom recovery logic required. Catalyst 2.0 extends this support across 10-plus frameworks, giving teams one consistent layer of reliability and trust regardless of which framework they have already standardized on. Catalyst 2.0 can run in the cloud, on-premise and fully air-gapped.
“The first wave of AI focused on making models intelligent, but the next wave is all about making AI systems trustworthy,” said Yaron Schneider, co-founder and CTO of Diagrid and chair of the Agentic AI Foundation Workflows Working Group. “When an AI agent triggers a tool, calls a service, or hands work to another agent, organizations need proof of what happened and confidence it will finish the job. Catalyst 2.0 brings that durability and verifiability to the frameworks teams are already building on.”
Durability and Trust for Every Major AI Framework
Catalyst gives developers a single way to harden agentic systems across existing production frameworks, rather than forcing a switch to a new framework or a custom-built reliability layer. The release adds support for:
- LangGraph and LangGraph Deep Agents
- Microsoft Agent Framework
- Google Agent Development Kit (ADK)
- AWS Strands
- OpenAI Agents SDK
- Claude Managed Agents
- CrewAI
- Pydantic AI Agents
- Dapr Agents
Each integration gets durable workflows that automatically resume from the point of interruption, plus the verifiable execution capabilities recently introduced in Dapr 1.18: cryptographic history signing, execution lineage propagation and workflow attestation. Bottom line: security teams can confirm an agent did what it claimed to do; compliance teams get a chain of custody for every decision; and organizations operating under regulatory requirements can prove execution was completed, trace its origin, and confirm it was not altered. This is now possible on any cloud or hosting environment they choose, from hyperscalers to offline edge deployments.
“Every AI agent framework released in the past two years has made it easier to build agents, but none have made it easier to trust them in production,” said Mark Fussell, CEO and co-founder, Diagrid. “With Catalyst 2.0, developers keep the framework they’ve already chosen, and they gain the durability and verifiability that production AI systems require. That combination is what turns a prototype into something an enterprise can actually run.”
“In a rapidly evolving landscape of AI models and frameworks, Catalyst provides the stable foundation we can rely on,” said Wendelin Niesl, Head of End-to-End Core Application Engineering at ZEISS Group. “It enables us to build a sustainable, durable, and resilient platform for both AI and traditional workloads, combining the simplicity of a platform approach with the flexibility required by developers and operators alike. Catalyst helps us deliver solutions that are secure, reliable, and built for the long term — exactly what enterprises need to innovate with confidence.”
Diagrid Catalyst 2.0 is available today. Developers can learn more and get started adding durable and verifiable execution to their AI agents at https://www.diagrid.io/catalyst. To learn how Diagrid Catalyst delivers agentic durable execution, read Diagrid’s latest blog post and join our upcoming webinar.
About Diagrid
Diagrid helps enterprises build reliable and secure distributed systems and AI agents. Founded by the creators of the open-source Distributed Application Runtime (Dapr), Diagrid provides the agentic infrastructure layer that delivers durable execution, secure service‑to‑service communication, AI orchestration, and cryptographic trust for modern cloud‑native and AI applications. Its flagship product, Diagrid Catalyst, combines durable workflows with advanced identity, access controls, and zero‑trust style governance so teams can run long‑lived, mission‑critical agent workloads without losing state or compromising security. Diagrid is a leading maintainer of Dapr within the Cloud Native Computing Foundation (CNCF) and a member of the Agentic AI Foundation (AAIF).
Organizations in highly regulated sectors such as financial services and healthcare use Diagrid to orchestrate multi‑agent workflows, enforce fine‑grained data access, and cryptographically attest which agents did what and when across their infrastructure. Learn more at https://www.diagrid.io/.
View source version on businesswire.com: https://www.businesswire.com/news/home/20260728284090/en/
Media gallery
