Introducing WRITER Agent: A new agent design pattern for enterprise scale

By 2028, more than one-third of all business decisions will be made by AI agents. But reaching that level from the current baseline will require an order-of-magnitude increase in AI usage, and that kind of adoption won’t happen without a fundamental shift in how enterprises use AI.

Most agents today follow a consumer design pattern – built for individual convenience, loosely structured tasks, and one-off generations. They’re optimized for personal productivity, not organizational impact. Enterprises operate on a completely different axis: repeatability, accountability, system connectivity, and governance. Leaders need AI that delivers consistent, high-quality, compliant work across hundreds or thousands of people.

That’s why we’ve spent the last few months – both in the field and in the lab – developing a new agent design pattern purpose-built for enterprise complexity and scale. It shifts agents from “help me draft something” to “help my entire organization produce consistently high-quality work using our data, our processes, and our guardrails.” Most importantly, it democratizes AI agent creation so that the teams closest to the work build and manage the workflows they run every day.

Introducing WRITER Agent

WRITER Agent, available today, is a single intelligent interface for scaling enterprise work. Starting today, it is now the default experience on our platform. On the surface is a simplified, intuitive chat interface; underneath is a powerful agentic system that introduces several new capabilities designed to support the way business users actually work:

• Playbooks for creating repeatable, natural language workflows
• Routines for running those workflows automatically
• Connectors that integrate with your systems and data
• Personality profiles that infuse your brand into every output
• Enterprise controls that govern web, data, and system access

At the core of all of this is a powerful reasoning engine powered by Palmyra X5, which combines robust, systematic planning with a comprehensive toolkit for handing virtually any type of enterprise workflow, from deep research and content generation to data analysis, code generation, and more. 

Built to fit your workflows

We purpose-built WRITER Agent for the knowledge work that define modern enterprises. Every team has workflows they repeat again and again, but never in the exact same way: responsibilities like persona research, market analysis, content creation, forecasting, and board prep. These aren’t one-off tasks – they’re recurring workflows that demand consistent methods at scale, but still depend on human judgement each time they’re run. 

In most organizations, these processes live in tribal knowledge, scattered documents, or a patchwork of manual steps. They’re difficult to standardize, hard to operationalize, and nearly impossible to scale without significant technical lift.

Playbooks

Playbooks change this paradigm entirely. They turn recurring tasks and processes into natural-language templates that alow you to repeat and scale work across teams. Purpose-built for business teams, playbooks are:

• Intuitive: Anyone can set them up – no technical expertise required.
• Contextual: Incorporate dynamic variables, connect to third-party systems, and adapt to the unique needs of different teams and roles.
• Effortless: Run them on demand or schedule them as recurring routines.
• Powerful: Built to run in Agent mode, they use structured planning and tool use to handle complex, multi-step workflows.

And soon, you’ll be able to configure routines to trigger automatically based on events or signals, such as receiving an email from a specific account or a status change in your CRM.

Connectors

A lack of access to enterprise context is what prevents most AI agents from actually driving value. Connectors, available today in open beta, access the tools, data, and systems your teams use every day, allowing WRITER Agent to go beyond just assisting with tasks to acting as a true operator within your ecosystem. We built these pre-configured, governed integrations on our own Model Context Protocol (MCP) gateway, developed in partnership with leaders like Snowflake, Microsoft, Hubspot, and Google, connectors are the foundation of true agentic orchestration.

Business teams can use playbooks, routines, and connectors in countless ways, such as:

• Account prep: Pull upcoming meetings from your Google or Outlook calendar, enrich them with context from your CRM, and generate complete, personalized briefing documents. 
• Campaign insights: Pull in and analyze campaign performance data to uncover engagement trends, then translate findings into a targeted email nurture sequence.
• Derivative content: Transform a webinar transcript into landing page copy, a blog post, a social media plan, and slides to add to a sales deck.
• BDR outreach sequencing: Create a personalized BDR email sequence that researches a prospect and tailors the value props to their needs.
• Investment analysis: Research a target company and create an investor presentation, including forward outlook with investment considerations.
• Inventory risk analysis: Use inventory data to identify SKUs with high stock, low velocity, or excessive discounting by size or category and deliver insights in a report.

Together, these new capabilities unlock new levels of automation – opening the door to new, high-value use cases across teams and verticals.

Personality profiles

Maintaining a consistent voice across different personas, channels, and sub-brands isn’t just hard – it’s costly and resource-intensive. Inconsistencies in the way your brand is portrayed affect everything from customer experience to regulatory compliance. Most AI tools make this problem worse: they generate content quickly, but rarely in a way that reflects a company’s unique voice, messaging, or standards. 

With personality profiles, WRITER Agent allows you to create and select voice profiles that carry through every type of output generated – from research briefs and competitive analyses to sales collateral, customer communications, and all forms of content – ensuring every output sounds like you, not a generic AI.

Presentation generator

Creating presentations is still one of the biggest time sinks for enterprise teams. Hours go into formatting formatting layouts, aligning visuals, and rewriting copy – work that can take even longer without design expertise. Current AI tools rarely help, producing generic slides that lack substance and accuracy. 

WRITER Agent’s presentation generator changes the equation. From board materials and sales decks to research summaries and strategy presentations, WRITER Agent handles the heavy lifting – structuring the narrative, integrating your enterprise context, producing charts and visuals, and infusing your brand – so that teams can go from idea to presentation-ready in minutes.

Enterprise-grade controls

Moving toward an AI-native operating model means enterprises will increasingly delegate decisions to agents. And for agents to make the right decisions, they need access to more data, more systems, and more tools. There’s a tradeoff: as agent capabilities expand, so does the surface area for risk.

To support safe and scalable adoption, the WRITER platform provides a comprehensive set of governance controls that let enterprises define exactly how agents operate:

• Web and data controls for defining permitted and restricted website lists, managing access to external sources, and applying granular data boundaries so agents operate only within approved contexts. 
• Third-party connector controls for governing which users or teams can access specific connectors, and for specifying exactly which tools (actions) within each connector are allowed.
• Agent visibility and auditability for tracking usage, monitoring flagged sessions, auditing compliance, and ensuring accountability across large teams.

What’s next: The AI-native enterprise

WRITER Agent represents our vision for how enterprises will become AI-native: by giving business users both the freedom to do their best creative and strategic work, and the power to scale it across their teams. The people closest to the work should be the ones shaping how AI supports it, and with the right tools, they can build the agentic systems that will take their organizations AI-native.

This frees up engineers to focus on the infrastructure that makes it all secure, scalable, and governed. When tens of thousands – or hundreds of thousands – of autonomous agents are executing work across every corner of your org, governance becomes existential.

So how does IT empower business teams to innovate, but do it safely and securely? We’re working on that too. Stay tuned.