Skip to main content

Introduction to Maz

A technical overview of Maz, Ninety's AI companion.

Written by Tommy Mains

What Is Maz and How Does It Work

Maz is Ninety's AI-powered set of features that uses specific, guided prompts with a large language model (LLM) along with user input and data in Ninety to complete a task. When using Maz, it has access to the same level of information in Ninety that you do: the same teams and permissions.

We designed Maz to both make it easier to run your organization's operating system and to meaningfully improve your workflows. Instead of functioning like one of our other tools, Maz uses a number of skills to analyze, create, and improve your data in Ninety.

Most of the time, you'll work with Maz directly. Open it from the Maz button on any page to ask about your data across every team you can access, understand Ninety and EOS methodology, and have Maz create and update items such as Rocks, To-Dos, Issues, Headlines, Measurables, and Scorecard scores. Read Working with Maz, Ninety's AI Companion to learn how.

Maz also includes these skills:

  • Personal preferences: Tell Maz how you want it to respond, such as a nickname, a reply length, or a language, and it remembers that preference in every future conversation. Read Teaching Maz Your Preferences for more information.

  • Seat Creator/Refiner: Create a new Seat on your Accountability Chart based on EOS principles and your company's information. Read Creating and Refining Seats with AI-Assistance from Maz for more information.

  • SMART Rock Creator/Refiner: Create a new Rock (90-day goal) using EOS's interpretation of the SMART methodology — or refine an existing Rock from its details panel — by answering a series of questions from Maz. Read Creating and Refining Rocks with AI Assistance from Maz for more information.

  • Scorecard Optimizer: Analyze your team's weekly Scorecard with Maz using EOS best practices and then decide to implement Maz's suggestions for changes. Read Optimizing Your Scorecard with Maz for more information.

  • Meeting Prep Emails: A personalized summary of what to review before your Level 10 Meeting®. Read Meeting Prep Emails from Maz for more information.

Our product and engineering teams are in the discovery, design, and production phases for additional Maz skills. To stay up to date with new releases, visit our Product Updates page. Also, consider joining our beta program for early access to many of our new features.

Using Maz

The top of the My 90 page in Ninety. A green box highlights the Maz button in the upper-right corner of the page.

Click the Maz button in the top-right corner of any page in Ninety to open Maz. If you minimize a conversation, a blue dot appears on the button; click Maz again to resume where you left off. To learn how to ask questions, have Maz take action, and manage your conversations, see Working with Maz, Ninety's AI Companion. To revisit an earlier conversation, see Viewing and Continuing Past Maz Chats.

Other Maz functions, such as the Seat Creator/Refiner, Scorecard Optimizer, and SMART Rock Creator, are launched from their respective tool pages: the Accountability Chart, Scorecard, and Rocks pages. You can also refine an existing Rock with Maz from the Rock's details panel.

Which LLM does Maz use?

Our team tests various generative AI models for each Maz skill, then selects the one that performs the most effectively and efficiently. Maz uses the selected LLM to process pre-set prompts, user input, and data from your company's Ninety account. Our vendors, including web service providers and AI companies, cannot access your prompts, Maz's responses, or any data sent between them.

Your data is never shared with model providers and is never used to train/improve base models. Read our Privacy Policy for more information.

How Maz connects to your data

Each Maz skill listed above uses an MCP layer to access a set of "tools" using Ninety's private and secure API (Application Programming Interface) pathways to read the specific information it needs to complete its task.

What is MCP? MCP, or Model Context Protocol, is an open standard for securely connecting generative AI models to systems, such as your Issues list in Ninety.

For example, when using the Seat Creator skill, Maz may use the pathways in our MCP/API infrastructure to access and read your organization's Directory tool to view user names in order to accurately assign the Seat's owner.

Who can use Maz and what it can access

Maz is available to all users and companies on any paid subscription plan.

Maz provides different levels of access to your company's data based on each person's Ninety account, including their user role and the teams they're assigned to. Only licensed users (Owners, Admins, Implementers, Managers, and Managees) can create and edit items in Ninety, so only users with those user roles can use Maz to create and edit items in Ninety. Preferences you save with Maz change only how it responds to you; they never change which data it can access.

Each team created in Ninety has its own Issues page, Scorecard, and To-Dos list. However, Owners and Admins can access every non-private team in your Ninety account.

For example, if your leadership team's V/TO is not shared with other teams in your organization, users not assigned to the leadership team cannot ask Maz to analyze the leadership team's 1-Year Plan, Core Values, Company Rocks, and so on.

For a role-by-role breakdown, see the Who can use Maz section of Working with Maz, Ninety's AI Companion.

Why is it called "Maz"?

Pyramid diagram of Maslow's Hierarchy of Needs with five tiers, from bottom to top: Physiological, Safety, Love/Belonging, Esteem, and Self-Actualization.

Maz's name is inspired by Abraham Maslow, an acclaimed American psychologist. Maslow introduced the "Hierarchy of Needs" in 1943. Ninety's founder, Mark Abbott, has always seen it as a helpful framework for understanding how people can achieve their highest potential.

During Ninety's initial development, Mark began planning how AI would eventually be used to supplement the platform and provide value to our customers. We championed the name "Maz" as a reminder that our AI features should help the people doing the work at their organizations reach their potential.

Did this answer your question?