Introducing agent supervision designed for enterprise scale

Today, we’re launching our agent supervision suite — giving IT leaders the visibility, control, and confidence to govern agents at enterprise scale. These new features are available in early access, with availability expanding over the coming weeks.

The timing couldn’t be more critical. According to Gartner, 50% of all knowledge workers will work with, govern, or build their own agents within two years. We’re witnessing the mass democratization of AI development‌ — ‌and WRITER is leading this movement.

We put powerful capabilities into the hands of non-technical business users, enabling them to create, adopt, and scale AI agents for any mode of work, from structured tasks to autonomous flows. The people closest to the work can innovate on myriad use cases, large and small — freeing technical teams to focus their resources on building complex use cases and adapting the best ideas from the business.

There’s a tension I hear about from every technology leader I talk to — on one hand, they’re drowning in backlogs of requests from business users, which makes it harder to focus on the really complex, high-value projects. They’d welcome the chance to let the business build and test new ideas, proving their value before IT expends resources.

But on the other hand, they’re concerned about what it means to democratize access to building agents. The very autonomy that makes agents so powerful also makes them a significant liability if left unchecked.

The question becomes urgent — when business teams can build at scale, how does IT create the conditions for secure AI adoption? How do you govern all of these increasingly powerful agents cropping up across the enterprise?

That’s exactly what we’re solving with this new feature set. Building on WRITER’s AI-native foundation of trust and security, we’re introducing enterprise-grade governance that gives IT teams the tools to observe, control, and connect every aspect of your WRITER deployment. It’s the infrastructure that makes democratized AI development actually viable at enterprise scale.

Get full visibility across your WRITER deployment

The challenge: Business teams are set to deploy dozens of agents in the next quarter. But which ones are actually being used? Which ones are accessing sensitive data? Which ones are quietly burning through your AI budget? Without centralized visibility, you’re flying blind‌ — ‌unable to optimize performance, manage costs, or ensure compliance.

The solution: WRITER’s comprehensive monitoring and analytics provide a single pane of glass into every agent, every interaction, and every cost across your deployment.

Admins will be able to:

• Track usage and performance across the organization, with both aggregate trends and user-level activity.
• Review event details, including request-level metadata, to audit agent behavior and compliance.
• Manage costs proactively by setting and monitoring rate limits centrally, with automatic query caching to reduce spend.
• Identify and resolve issues before they impact users, with performance monitoring and customized alerts across agents and knowledge sources.

Monitoring data integrates with your existing observability stack, ensuring centralized visibility without adding another tool to manage‌ it.

The result: With centralized visibility, IT teams can quickly identify performance bottlenecks, improve underperforming agents, and prevent cost overruns before they impact budgets. Instead of reactive troubleshooting, you can proactively manage your entire AI deployment from a single dashboard.

Enforce policies, protect data, and deploy with confidence

The challenge: Visibility alone isn’t enough. You need the ability to take action — preventing problematic agents from deploying, enforcing security policies at runtime, and making sure that every AI-powered interaction meets your requirements for compliance and data protection.

The solution: WRITER makes it easy to govern the entire agent lifecycle, from development through deployment and operation.

IT teams gain active control with:

• Agent controls: See every agent being built in WRITER. Review and approve agents before they are deployed, preventing shadow AI from proliferating across the organization.
• Global policies and guardrails: Set system-wide policies once, and configure guardrails that automatically block or mask sensitive data across all agent and user actions. No need to configure each agent individually.
• Granular permissions: Customize role-based permissions that are applied at runtime across agents and connectors. Define clear responsibilities and provide least-privileged access to users — ensuring that agents can only access the data and systems they need, nothing more.

Our approach is to provide fine-grained controls, implemented globally — guaranteeing that your AI workflows are secure and compliant from the ground up, and scale automatically as adoption grows.

The result: IT teams can confidently scale agent adoption while maintaining control. Approve the agents that meet your standards, block those that don’t, and make sure that every deployed agent operates within the guardrails you’ve defined‌ — ‌automatically.

Connect WRITER with your existing tech stack

The challenge: Your IT teams already rely on established platforms for observability, security, and guardrails. The last thing you need is another siloed tool that doesn’t integrate with your existing infrastructure‌ — ‌forcing you to manage AI governance separately from everything else.

The solution: WRITER is open by design, integrating natively with the tools your teams already use. Two weeks ago, we announced WRITER Connectors to securely connect agents to enterprise apps. Now, we’re making WRITER interoperable with the platforms IT relies on:

• Observability platforms like Datadog and Traceloop
• Security platforms like Noma and Lakera
• Guardrails providers like Amazon Bedrock Guardrails and Lasso

These prebuilt integrations allow you to improve oversight and maintain consistency across all of your AI deployments — managing WRITER agents through the same tools and workflows you use for everything else.

We’re building an ecosystem of established providers and AI-native platforms — if we don’t integrate with your preferred platform yet, we want to hear about it.

A deeper partnership with AWS

This morning at AWS re:Invent, our CEO, May Habib, announced a deepened partnership with AWS‌ — ‌one that fundamentally expands what’s possible with enterprise AI governance.

What’s new: You can now use third-party models — including those you’ve custom-trained or fine-tuned‌ — ‌to power agents in WRITER, via Amazon Bedrock. We’re also integrating Amazon Bedrock Guardrails, so admins can configure guardrails for WRITER agents using the same data protections and policies deployed elsewhere in their AI stack.

Why this matters: With universal model controls, you have the flexibility to choose the best tool for the job while managing everything through a single, unified governance layer. This means full confidence in your model choices, routing decisions, and compliance enforcement‌ — ‌without sacrificing the interoperability enterprises demand.

Why Amazon Bedrock? Our customers trust WRITER to meet the highest standards for enterprise-grade performance, data privacy, and security. In expanding our platform to support external models, we knew we needed to do it in a way that maintained the enterprise-grade standard we and our customers expect. Amazon Bedrock met these standards and had already earned the trust of many of our mutual customers‌ — ‌making them the perfect launch partner.

Building the future together

These new supervision capabilities represent a critical milestone for enterprise AI — ‌solving the governance challenge that has held back widespread agent adoption. For the first time, enterprises can confidently democratize AI development while maintaining the control and security that Global 2000 standards demand.

Our expanded partnership with AWS reinforces what’s becoming clear — the future of enterprise AI requires platforms purpose-built for both innovation and governance. The days of choosing between empowering teams and maintaining control are over.

As technology leaders navigate this transformation‌ — ‌from managing systems to governing intelligent agents working alongside every employee — WRITER is committed to partnering with you every step of the way. The companies that will lead in this new era are those building the infrastructure for supervised autonomy today.

If you’re navigating how to scale AI adoption while maintaining governance, or defining what your infrastructure needs to look like when agents outnumber employees, we’d love to partner with you.

Learn more about this release. Request early access.