主なポイント
- Multiquip built a customer-facing AI parts agent with Boomi Agentstudio that finds the right item from 200,000 SKUs in seconds, down from lookups that could take 10 minutes or more.
- Connecting previously siloed systems — Documoto, SAP, SmartEquip, Snowflake, and Azure — lets the agent answer questions that used to require several systems and a subject-matter expert.
- Multiquip expects a 90% cut in parts lookup time, three to four times more support productivity, and a 3% lift in parts revenue, while freeing technicians for higher-value selling.
Most of us have experienced wandering aimlessly down the aisles of a big-box hardware store because a well-meaning associate couldn’t locate the right part for our do-it-yourself project. Multiquip, Inc. refuses to replicate that frustration for its professional customers in the high-stakes, high-pressure world of construction and portable prime power generation.
One of the world’s most diversified suppliers of machinery, Multiquip views customer satisfaction as non-negotiable. Downtime is more than a delay in construction and manufacturing. It’s a crisis that drains the bottom line. When equipment fails, jobsites go silent, power generation stops. Contractors quickly need replacement components fast — and right — every single time. There’s no margin for error in the hard-hat industry.
This is why Multiquip used Boomi Agentstudio to build a production-grade, customer-facing AI parts order agent: keeping customers working. The agent interprets natural language queries, understands industry jargon, and retrieves accurate parts data for 700-plus pieces of equipment across multiple enterprise systems, with diagrams, schematics and additional information — all through a single conversational interface. In addition to speeding up the order process, the agent boosts revenue by recommending items the customer might also need while giving parts specialists additional time for outbound selling.
For a mid-sized company, this AI-driven innovation is a great market equalizer, leveling the playing field against larger rivals.
“Being able to scale this up and leapfrog some bigger competitors who started earlier and with bigger teams is a huge advantage for us,” said Michael Hanken, senior vice president of IT at Multiquip. “That’s the classic mid-market approach. Same complexity, less people and tools.”
Here’s a closer look at the role Boomi plays in orchestrating Multiquip’s AI journey and helping keep jobsites active.
Disrupting a Traditional Industry
Hanken knows it sounds unusual that a construction parts business is at the forefront of agentic adoption. But, he added, of all the risks associated with AI, the biggest is to do nothing.
“We are not an innovative industry by trade,” he said. “Construction ranks second to last in terms of innovation aptitude, and the worst is forestry. Construction has its own language. The people are down-to-earth, and it’s amazing what they do. But when a machine is not working, nobody is working. They need it fixed right away.”
Multiquip has relied on Boomi’s B2B/EDI capabilities for integrating key business operations since 2022. That trust helps explain why Hanken, who recognized AI’s business potential early, was interested to hear Boomi’s perspective. Multiquip leaders took part in a day-long AI-readiness workshop hosted by Boomi’s Head of Research and Emerging Technology, Michael Bachman. In that pre-agent AI era, they settled upon a use case involving customer access to repair manuals and worked with Boomi on a trial. But the feedback from Multiquip’s demanding customer base was underwhelming, and it didn’t look like the project would move the business needle.
Hanken said they learned invaluable lessons from that prototype, particularly about ensuring that accurate data was available to avoid hallucinations — wrong information for customers. Undertaking a project to make Multiquip’s vast manual catalog more AI-ready would pay big dividends as rapid advances in agents were occurring.
“There’s not much room for error when you do something customer-facing,” he said. “I’m a process guy because that’s what I’ve done for 30 years, starting at SAP. Processes are important because they are what customers work with. We can have the niftiest bot in the world, but it’s useless if customers don’t use it. As a former CEO always told me, ‘Michael, you don’t create a solution where you need to chase after a problem.’ So we got very pragmatic.”
As Multiquip went back to the drawing board, Boomi released Agentstudio for the design, implementation, and management of agentic processes. At the same time, Multiquip’s vice president of customer success approached Hanken with a blunt request.
“He said: ‘Help me to sell more parts,’” Hanken recalled. “That’s when our wheels started turning.”
Staggering Complexity
Conversations with desperate contractors out in the field, urgently needing key pieces of machinery repaired, are stress-filled. Finding the right spare part with any service vendor can be an adventure, but the complexity of Multiquip’s parts catalog is mind-numbing.
The company has 200,000 different SKUs. As a result, taking orders was a manual, “swivel chair” process handled by a team of 10 service technicians. Every call involved back-and-forth questions as support specialists identified the needed replacement parts from over 1,000 equipment manuals and hundreds of PDFs across multiple, disconnected enterprise data systems.
The average lookup time could take 10 minutes or more per inquiry. Some errors were inevitable, and the process was impossible to scale for a growing team that requires deep product knowledge. It was not an ideal process for impatient customers, who require instant, accurate service to get machinery up and running.
It was here that Hanken saw a strong use case.
“It was low-hanging fruit where the agent can serve two purposes,” he said. “No. 1, the customer spends less time ordering the part. No. 2, we can repurpose our technicians to be more helpful selling parts on an outbound basis, increasing revenue.”
Multiquip worked with the Boomi AI Center of Excellence (AI CoE) and the Boomi product team to help Boomi Platinum Partner Jade Global develop the agent using Boomi Agentstudio. Its power derives from Boomi’s ability to connect what previously couldn’t be integrated easily in Multiquip’s tech stack: a complex network of systems that includes the Documoto parts catalog solution, SAP ERP, SmartEquip ecommerce, Snowflake, and Microsoft Azure Blob Storage. That connectivity empowers the agent to accurately answer questions that had required multiple systems and a subject-matter expert.
Converting existing integration processes into agent tools rather than building them from scratch reduced development time by an estimated 50%. Added Hanken: “We started this project last September, and we first rolled out a litmus test in February. So, the development was very rapid.”
How The Agent Works
Now, there’s no need for a phone call. A customer accesses the self-service agent online and provides details such as model and serial numbers. The agent autonomously retrieves the part information and provides a confirmation diagram. This happens all in a matter of seconds, significantly reducing ordering time. Also, the agent makes recommendations for related parts and services the customer might need, potentially increasing the purchase amount.
“With Boomi, our hallucination rate is less than half a percent,” Hanken said. “Fuzzy answers are not acceptable for our use case. It’s yes or no. There’s nothing in between. We can’t ship the wrong part. Customers have no patience for that.”
Only if the agent can’t resolve the issue are customers routed to a human support specialist. As the agent went into full production, Multiquip expected:
- A swifter response time vs. traditional phone calls (between five and 10 times faster), translating to a 90% reduction in parts lookup time, and a three-to-four times increase in support productivity.
- Customer service technicians with deep product expertise could focus on outbound customer calls to explore upsell and cross-sell opportunities, once the workload on more mundane inbound calls was reduced.
- A 3% increase in parts revenue annually.
“In the testing phase, one of our best technicians entered a question into the bot,” Hanken said. “He said it was much faster than him looking it up”. That told us we were on the right track. The technicians also say that the customers they talk to will be very happy.”
Multiquip recently presented the AI agent at two industry conferences, creating buzz as one of the first equipment distributors to achieve measurable agentic success and demonstrating what’s possible even in their tradition-bound industry. In fact, representatives from two big box chains were interested in learning more about how Multiquip did it, Hanken added.
Then there’s the competitive advantage the company has gained. Hanken said he understands Multiquip has gone to market much faster than one direct rival, which is larger and has greater brand visibility. The difference is that the company is designing all its AI processes internally.
“They have to build everything end-to-end,” he added. “We use Boomi.”
FAQ
What is Boomi Agentstudio?
Boomi Agentstudio is a no-code platform for designing, building, and managing agentic AI processes, letting teams turn existing integrations into tools an AI agent can use.
How does Multiquip’s parts order agent work?
Customers enter details like a model or serial number into a self-service, conversational agent. The agent retrieves the correct part, shows a confirmation diagram, recommends related parts, and completes the transaction in seconds. Requests it can’t resolve are routed to a human specialist.
What systems does the agent connect to?
The agent connects Documoto’s parts catalog, SAP ERP, SmartEquip ecommerce, Snowflake, and Microsoft Azure Blob Storage — systems that previously required manual, cross-referenced lookups by service technicians.
What results has Multiquip seen from the AI agent?
Multiquip expects a 90% reduction in parts lookup time, a response five to 10 times faster than phone calls, a three-to-four times increase in support productivity, and a projected 3% increase in parts revenue.
How did Multiquip reduce development time?
By converting existing integration processes into agent tools rather than building them from scratch, Multiquip and Boomi Platinum Partner Jade Global cut development time by an estimated 50%.