Many Requests
Deploy AI teammates that join meetings, learn your business, and help teams move faster.
Artificial Intelligence
Product Strategy & AI Experience Design
Role
Lead Product Designer
Timeline
12 Weeks
team
AI Workflows Product Strategy Information Architecture Workflow Design Design Systems
platform
Web
The Observation
Most companies don’t struggle because they lack information.
If anything, they have too much of it.
Customer conversations happen every day. Decisions are made in meetings. Teams share updates in Slack. Processes get documented, revised, and passed between people. Over time, all of that knowledge starts living in different places, owned by different people, across different tools.
The information still exists, but finding it becomes a job of its own.
What Made This Interesting
What interested me about Canopy was its approach to solving that problem.
Rather than building another AI chatbot, the product explored a different idea: what if organizations could create specialized AI teammates that actively participated in the way teams already work?
These agents weren’t designed to simply listen and take notes. They could join meetings, participate in sales calls, access company knowledge, answer questions, and communicate on behalf of the organization when appropriate. Depending on their role, they could engage in conversations, provide information in real time, and help move discussions forward.
The more we explored the product, the more questions started to emerge. If a company has multiple AI teammates, how do people know what each one does? What information can they access? When should they step in? And how do you make all of that feel natural instead of overwhelming?
The Solution
Canopy was designed as a platform where organizations could create and manage specialized AI teammates across different functions of the business.
Instead of relying on a single assistant, teams could deploy agents with specific responsibilities, permissions, and knowledge sources.
Some agents focused on sales and customer conversations.
Others supported onboarding, operations, internal support, or company knowledge.
These agents could:
(1) Join meetings and actively participate in discussions
(2) Attend sales calls and answer product-related questions
(3) Participate in Slack conversations across teams
(4) Access company documentation and internal knowledge
(5) Capture decisions, action items, and key insights
(6) Respond to questions using approved company information
(7) Support employees, customers, and stakeholders in real time
What emerged was less of a traditional software tool and more of a digital workforce designed to operate alongside the people behind the business.
What Changed
Before Canopy, important information was scattered across meetings, emails, chat threads, recordings, and documentation.
After Canopy, conversations became searchable. Decisions became easier to revisit. Knowledge became easier to share across teams.
Instead of depending on individuals to remember where information lived, organizations had a system that could retain context, surface knowledge when needed, and support conversations as they happened.
The product didn’t just help teams find information faster. It helped them spend less time managing knowledge and more time acting on it.
Why It Matters
As companies grow, context becomes harder to retain.
A customer shares something important during a sales call. A decision gets made during a meeting. Someone documents a process. A few weeks later, that information might as well not exist.
Not because it disappeared, but because nobody knows where to find it.
Teams spend time searching through Slack channels, meeting recordings, CRM notes, and internal documentation. Valuable knowledge gets buried. Conversations get repeated. Decisions get revisited. The larger the organization becomes, the more expensive that problem gets.
What Changed
One thing I kept coming back to while working on Canopy was that this wasn’t really a project about AI.
AI just happened to be the technology.
What made the problem interesting was how people work together. How knowledge moves through an organization. How context gets lost. How decisions get shared. And how quickly information becomes difficult to access as teams grow.
In many ways, Canopy felt less like designing software and more like designing how humans and AI could work together. Those are the kinds of problems I find myself enjoying the most.
