Zoho has announced updates to its Catalyst by Zoho Platform-as-a-Service (PaaS) to make it easier to move trusted AI-generated code to production. Code generation is no longer the measure of effectiveness in development teams. AI has seen to that. Now it is about how that code moves from generation to production and whether it can be trusted. That challenge has always been there but has been masked by how code is written.
Other solutions address this at the testing and code quality level. Catalyst is addressing the more complex challenge of integration. It will free up developers to use the AI coding assistant they want and provide a full-stack serverless model to deploy on. There are new Agent Skills, a new Command-Line Interface (CLI) and greater Model Context Protocol (MCP) support.
Sachin Agrawal, Managing Director of Zoho UK, said, “By delivering an agent-ready, serverless full-stack platform, Catalyst gives developers and AI agents complete cloud independence to build, deploy, and scale intelligent applications, with far less operational overhead.
“Additionally, to democratise access we are launching a free student tier, freeing students from the burden of traditional development cycles so they can thrive in a rapidly shifting landscape.”
What is Zoho adding to Catalyst?
There are four elements that Zoho is introducing into Catalyst to bridge the gap between code generation and reliable deployment.
Agent Skills provide context. They teach the AI assistant how to use Catalyst services correctly. Instead of guessing, the assistant follows recommended development patterns. This reduces token usage and improves accuracy, even when using lighter models.
The non-interactive CLI allows agents to execute multi-step workflows. A developer can initiate a deployment sequence that runs from start to finish without manual input. This accelerates the move from coding to testing.
The MCP server connects the AI assistant directly to Catalyst capabilities. Developers can create database tables or modify columns directly from their IDE. This keeps the workflow inside tools like VS Code or Cursor. It removes the need to switch contexts to a separate cloud console.
Orchestration ties these elements together. The Skill routes requests down either the CLI or MCP path. It does not leave this choice to the model. This reduces the cognitive load on developers and lowers the risk of incorrect tool selection.
Governance as a competitive moat
Zoho has talked often about the need for governance across the entire AI lifecycle. Agrawal talks in this podcast about how Zoho has built that into its own products as a core competency.
With Catalyst, it is focusing that governance on the lack of context that AI coding tools have to build secure applications. It is shifting governance and controls into that code generation process and specifically the cloud rules required for production. This is where it sees risk, and how embedding governance directly into the connection layer removes it.
By allowing Claude and OpenAI to connect through the new universal MCP URL, there is now an audit trail. Every action the AI takes generates a log which tracks specific commands. They show when an agent created a table or modified a function. Those are areas that are often overlooked in how AI interacts with the environment.
This approach also decouples development from production. Additionally, it provides a solution to an existing problem inside developer tools. Poor controls allow developers to use live data for testing. With Catalyst, Zoho claims that the AI agent can only build in a sandbox and cannot touch live data.
When code is moved to production, there is still a sign-off process. While some view keeping the human-in-the-loop slows things down, there has to be a control process. This is not just about governance but business risk. It also meets the demands to audit AI behaviour from regulators and reduces the fear of “black box” errors.
Will this give Zoho a competitive advantage?
As with all software releases, customers want a clear indication of what they are going to gain. With Catalyst, this is all about Zoho’s structural approach to AI governance as shown above. So how does it compare to other platforms?
GitHub Copilot excels at code completion inside the editor but lacks native controls for cloud infrastructure. It does not prevent an agent from configuring a database insecurely and leaves it to the developer to manually review every change. This creates a bottleneck at scale.
AWS Bedrock and Azure AI Studio also have powerful AI tools. However, they require complex stitching of separate services for security and logging. That means IT Ops teams are dealing with multiple consoles or writing custom consoles. It increases the risk of configuration errors and doesn’t provide audit support.
For system integrators, Catalyst reduces the cost of compliance. It removes the need to bolt on security tools after the fact. The platform itself provides the control. This approach positions Zoho as a safer partner for regulated industries. The question remains: will enterprises prioritise this integrated safety over the flexibility of fragmented cloud stacks?
According to Shiva Varma, Senior Director Analyst at Gartner, “When agents operate autonomously, actions are executed at a scale and speed that can outpace human oversight.
“Because accountability for outcomes remains with the organization, this level requires the most rigorous governance, including continuous monitoring, enforced guardrails, rapid rollback mechanisms, circuit breakers that halt agent operation on threshold violations and clear ownership for agent behavior.”
Enterprise Times: What does this mean?
Software vendors are all rushing to address weaknesses in the software development lifecycle (SDLC) model for AI coding assistants. Much of that is focused on the mechanics and improving speed and quality.
Zoho is also addressing that but recognises that the one thing that is needed is governance. It is looking outside of the IT function and code generation to deal with risk and compliance. After high-profile issues with AI coding assistants deleting databases, this is a necessary move. It also addresses increasing demands from regulators.
One customer, Gavin Smith, Founder, Element Technology Services, said, “When we built ConfigPilot, and later ActionPilot, we made a deliberate choice to build them on Catalyst rather than stitch together our own infrastructure. Catalyst’s agent-ready approach, where an AI agent can write, deploy, and manage the underlying services just as readily as one of our engineers can, helped us get to first release faster.
“It also enabled us to adapt to real-time feedback into a shipped change quickly, rather than waiting on a quarterly release cycle. The same holds for ActionPilot: being built on Catalyst means we can keep reshaping how it connects and behaves without re-architecting the platform underneath it every time a customer asks for something new.”
Catalyst by Zoho is available now with a monthly free tier for developers to explore the platform. There are also £200 ($250) in free credits for users who want to go deeper over a six-month period. There is also a completely free for students, option.
