How Assembled cut knowledge retrieval time by hundreds of hours with Dust
B2B SaaSKnowledge ManagementCompany-Wide AI AdoptionKey Highlights
- 95% internal adoption across 120+ employees without extensive training
- Dust adoption spread across multiple teams, including marketing, sales, customer success, engineering, and product
- Hundreds of hours saved across search and documentation tasks
About Assembled
Founded in 2018 by former Stripe engineers, Assembled is a customer support platform that uniquely combines AI agents with workforce management so companies can orchestrate both automated and human support on a single system.
By early 2024, the company had grown to nearly 120 employees and was evolving at a rapid pace. That growth accelerated even further when Assembled made a pivotal shift: launching four new AI products—chat, voice, email, and a live-agent co-pilot—all at once.
This leap transformed Assembled from a focused workforce management vendor into a complex, multi-product AI platform. But with that transformation came a new set of internal challenges: more products, more information, more surface area for questions, and no unified place to keep track of it all.
As Cassandra Stumer, Product Marketing Manager, puts it: “The speed at which Assembled moves is insane. We became a multi-product platform almost overnight.”
Challenge: Hypergrowth created knowledge chaos
While the product launch propelled exponential growth, it also changed the company’s operating rhythm in a way that exposed several sticking points:
Information lived everywhere and nowhere
Assembled’s rapid product expansion meant every team needed more information, faster. But the company’s knowledge was spread across an overwhelming number of places: Google Drive, Slack, Notion, Linear, Cursor, Snowflake, and more.
Even though employees were individually strong at using AI, their approaches weren’t shared or standardized. “Everyone at Assembled is really good at using AI,” Cassandra says. “The problem was everyone was using different tools like ChatGPT, Claude, Cursor, Runway, and none of it was shared. Amazing things were happening individually, but it was all siloed.”
Search was broken—and it slowed everyone down
With knowledge scattered across dozens of tools, search became a critical bottleneck. Most systems simply weren't capable of retrieving information across so many sources.
This was especially a problem for new hires who didn’t have the historical context to understand where to find certain information. Cassandra recalls having the same problem when she joined the company: “I’d been here four or five months and people would say, ‘Oh, that’s in the customer tracker.’ And I had never seen this document before. No one even told me it existed.”
Without a reliable way to find information, onboarding slowed, work was duplicated, and employees frequently had to ask colleagues where documents lived, which created friction and frustration across teams.
Rapid updates made knowledge quickly obsolete
Assembled’s engineering team ships at an unusually fast pace, sometimes multiple times per day. Updates typically appeared in a “shipped” Slack channel, which quickly turned into a never-ending scroll of changes.
“We’re so fast-moving that it wasn’t even clear what was alpha, what was Generally Available, what was ready for customers. It was just: ‘It shipped.’ And then everyone tries to figure out what that means.” This made it nearly impossible for GTM, CS, and sales teams to stay up to date—let alone onboard new people or maintain documentation.
The need for a unified AI strategy
Different teams began experimenting with automation through tools like Relay, Zapier, or individual Claude MCP setups. But these approaches didn’t scale since each workflow lived inside someone’s personal account, dependent on their bandwidth and technical confidence.
“We knew we needed to invest as a company in internal AI because so much was happening so quickly. But it was unrealistic to expect busy people to set up their own Claude MCPs or connectors. We needed something everyone could use.”
Solution: A founder-led discovery that solved everything
Assembled’s CEO Ryan Wang knew Dust’s founder, Gabriel Hubert, from their time at Stripe. Ryan began experimenting with Dust inside a Slack channel, using it to test small internal workflows.
Cassandra came across those early experiments and instantly recognized the potential.
“It was just incredibly easy to use. You connect sources once, in one place, and the agents can read and write across everything. I’m not super technical, and it was still simple.”
Search became the wedge that accelerated adoption
The first company-wide unlock was simple but transformative: Dust gave employees one place to search across all their tools. With a single query, teams could pull information from:
- Notion
- Google Drive
- Slack
- Gong
- Salesforce
The impact was immediate.“Search was the gateway,” Cassandra says. “Once people realized they could actually find things, adoption took off.” For the first time, Assembled had a unified, intelligent search layer that sat on top of all its systems.
Teams quickly moved from users to builders
As employees embraced search, they began building specialized Dust agents for their own workflows:
- Marketing built competitive intelligence and content automation agents
- Sales created follow-up and account research agents
- Customer Success built Zendesk and product feedback agents
- Engineering added code-base search and SQL query agents
- Product automated changelog and release mapping tasks
“Once you’ve used a couple of agents, building your own feels obvious,” Cassandra says. “People went from users to builders almost overnight.”
To reinforce learning across teams, Assembled created an internal AI Collective, which is a monthly gathering where employees share agents they’ve built and collaboratively troubleshoot new ideas. Hackathons, working sessions, and async sharing quickly followed. Dust was no longer simply a tool—it became a shared internal platform and a driver of organizational learning.
Results: Search solved, hours reclaimed, culture shifted
The adoption of Dust at Assembled delivered immediate and measurable impact across the organization.
A single source of truth for a fast-moving company
Dust unified Assembled’s fragmented knowledge ecosystem. Employees no longer guess which system contains the latest information—they simply ask Dust.
“Just in terms of searching and finding company information alone, Dust has saved us hundreds of hours.”
A dramatic reduction in cross-team interruptions
Before Dust, teams constantly asked one another where documents lived or how to find the most up-to-date materials. Today, that workflow has changed entirely. “It used to be: ask marketing or CS where something is. Now it’s: ask Dust. Everyone can self-serve,” Cassandra says. That shift freed subject-matter experts from low-leverage questions and gave every employee access to the same level of insight.
Faster onboarding with fewer blind spots
New hires now have a single place to understand Assembled’s history, decisions, documentation, and product context, regardless of which tool originally housed the information. This means that the company can now scale sustainably to meet the rapidly growing demand for their product.
From knowledge chaos to an AI-powered operating system
What began as a simple search problem quickly became the foundation for a new way of working. Dust now serves as the connective tissue across Marketing, Sales, CS, Product, and Engineering—turning scattered knowledge into a living, accessible system.
As a result, teams no longer rely on tribal knowledge or institutional memory. They don’t waste time hunting for information, duplicating work, or interrupting colleagues just to find the latest version of a document. Instead, they now have a shared internal platform where they can build together, learn together, and standardize their workflows.
"This tool is incredibly important to my everyday workflow. And I don't know how I would work without it,” Cassandra says.
