How AceMQ turned 300 hours of consulting calls into a knowledge base that cut tickets 30%

AceMQ sells what its engineers know about RabbitMQ, Kafka and Redis. Before Documentation.AI, most of that knowledge sat in call recordings, engineering notes, Notion and Google Docs, and very little of it was ever written up properly.

Now it lives in one Documentation.AI workspace. The team writes into it through the Authoring MCP server, clients search it before they open a ticket, and AceMQ's AI assistants read it live through the Reader MCP server.

Results

About AceMQ

AceMQ is a messaging consultancy founded in 2017, headquartered in Miami and part of ace8 Technologies. It describes itself as Broadcom's exclusive strategic RabbitMQ managed service provider, and works with more than 130 enterprises in 26 countries on RabbitMQ, Kafka, Redis and VMware Tanzu, from consulting and migrations to training and round-the-clock support.

A firm like that is only as good as what its people remember. Every engagement teaches the team something about how a broker behaves under load, or how a particular client has set theirs up. The question was where that knowledge went once the call ended.

Before: everywhere except somewhere useful

Tyler Eastridge, AceMQ's COO, doesn't soften it. "Before Documentation.AI, our knowledge lived everywhere except somewhere useful."

Client technical details were buried in call recordings and engineering notes. SOPs were split across Notion, Google Docs and people's heads. There was no version control and no way to tell whether a page was current, so documents got written once and then left alone.

"Everything was either out of date or just dead," Tyler says. "Documents sitting in folders that nobody touched after they were written."

What finally pushed them was taking stock of what they were sitting on. "We had years of genuinely valuable proprietary knowledge with no way to organize, maintain, or monetize it. In the AI age, your proprietary data is your main competitive asset. Ours was scattered and inaccessible."

What AceMQ evaluated

AceMQ looked at documentation platforms and internal knowledge tools side by side, because it needed one system to do both jobs.

"They're built for internal notes, not for documentation as a living, organized product across multiple use cases," Tyler says of everything except Mintlify. What decided it was "the combination of beautiful UI/UX, structured content management, and the Authoring and Reader MCPs. No other platform we evaluated had all three."

The worry that nearly stopped it

AceMQ's clients run messaging systems they can't afford to have fail, and a wrong troubleshooting step in a client-facing article is a real problem. The team went in doubting that AI-assisted content would be accurate enough to publish.

"That concern went away quickly once we saw the authoring workflow in practice," Tyler says. The AI drafts from the team's own session notes, and nothing goes live until someone on the team publishes it.

How the team writes now

The Authoring MCP server was the turning point. AceMQ connected it to the AI tools the team already works in, so a consultant can go from raw session notes to a structured, published page in the same sitting. Writing that used to take half a day takes minutes.

That is what finally cleared the backlog. More than 300 hours of consulting transcripts are now published and searchable, along with hundreds of delivery artifacts. None of it would have been written up by hand, because nobody was going to give up an afternoon per call to do it.

"For the first time ever, we have documentation that actually stays up to date," Tyler says. "That sounds simple, but it's genuinely the first time that's been true for us."

“In the AI age, your proprietary knowledge is your most valuable asset. Documentation.AI is the first platform that finally gave us a real knowledge management system, where we can write, organize and keep everything up to date in one place, with a UI that people actually want to use.”

One workspace, five audiences

All five draw on the same pages, so a correction made once shows up everywhere. In Tyler's words, "We went from disparate Google Docs and dead files to a living, breathing knowledge management system."

Fewer tickets, and still falling

Support tickets are down 30% since AceMQ opened its knowledge base to clients, and the count keeps dropping because of how the team handles the rest. When a new ticket comes in, the answer goes into the knowledge base, so the next client with the same question finds it without asking.

Inside the company, having everything in one place saves at least five hours a week. Consultants find past answers in seconds instead of searching old recordings, and new consultants learn from the same pages the senior team writes.

AI assistants that read the live docs

Through the Reader MCP server, AceMQ's internal tools and its client-facing AI assistants query the knowledge bases directly. They answer from the current page rather than an export from months ago. The site's access rules apply to them too, so a gated page stays gated however it's reached.

Selling the expertise

Once the knowledge was organized, current and readable by AI tools, AceMQ could sell it. The firm now packages its accumulated expertise as a premium offering that clients pay for, something it had no way to do while the same material was spread across recordings and folders.

Before and after

"Whether it's a SaaS doc site, an internal knowledge base, a client-facing KB, or an AI-connected knowledge layer your tools query live, it handles all of it," Tyler says. "It completely changed the game for us."