With AI On The Rise, One MSP Builds A Chatbot With A Cajun Accent

The CEOs of Phantom Technology Solutions and General Informatics talk AI adoption.

Meghan Landry, one of the more recent additions to General Informatics, has a Cajun accent, a habit of calling people “bon ami,” can rattle off the company’s paid time off policy in an instant and hails from Breaux Bridge, La., about 50 miles from the MSP’s Baton Rouge headquarters.

But Landry doesn’t eat, sleep or breathe like her human colleagues.

That’s because Landry is a chatbot – or rather, a persona built into a large language model (LLM) powered chatbot used internally by General Informatics, CEO Don Monisterem told a crowd of solution providers at CRN parent The Channel Company’s XChange NexGen 2023 conference, which runs through Tuesday in Houston.

“You can really do some interesting things by defining the persona side of the LLM,” he said. “And it will help actually answer the questions.”

[RELATED: AI And The Channel: It’s Go Time]

MSPs Look Into Chatbots, GenAI

Monisterem detailed the chatbot made by his company – a member of CRN’s 2023 MSP 500 – during a panel on the ongoing adoption of generative artificial intelligence (GenAI) by solution providers to improve internal operations and as a new offering for customers.

Tanaz Choudhury, CEO of Houston-based TanChes Global Management, another MSP500 member who watched the panel, told CRN in an interview that AI is here to stay.

Like any new technology, AI will have uses that are helpful to solution providers and their customers. And others will use AI for cyber attacks and other nefarious purposes.

“Ready or not, it’s happening,” she said.

A survey of 300 respondents conducted by The Channel Company showed that the promise of AI has captured the imagination of solution providers’ customers, with almost 50 percent of survey respondents saying they have received inquiries from customers about using GenAI or LLMs in their business.

About 11 percent of respondents in the survey said they already have full-scale AI and LLM implementations internally. Meanwhile, another 11 percent haven’t yet considered the technology for internal use.

Kristy Davis, global director of innovation and AI at CRN parent The Channel Company and another panelist, told the crowd to consider internal use cases such as using GenAI to author blog posts, marketing messages and sales presentations. MSPs should be thinking about business models around the technology.

The largest technology vendors have invested billions of dollars in upstarts in the space, but Davis predicts acquisitions on the horizon.

“You can see that there’s a big race here to get to market,” she said. “And what’s going to happen in that consolidation over the next year is going to be very interesting.”

She said customers could start adopting new AI technology within three years. “The market is going to move fast,” she said.

Another panelist, Henry Timm, CEO of Rolling Prairie, Ind.-based Phantom Technology Solutions – another MSP 500 member – told the crowd that his company has used AI internally to generate marketing materials. Employees at Phantom are working with clients on use cases for LLMs. He also sees MSPs working with open source offerings in the space on tools aimed at solution providers.

While GenAI has dominated headlines this year, Timm advocated taking a slower approach to adoption. Users need to consider data classification, biases and prompt data associated with LLMs.

“A lot of people unfortunately just started dumping all types of information, not sanitizing it, into ChatGPT when that came onto the scene,” he said. “Unfortunately, people aren’t giving thought to what that data is, who owns it, who has access to it. And that’s a risk you need to contemplate.”

Still, MSPs should leverage the knowledge leadership and “trusted adviser relationship” they have with customers to steer them toward safe experimentation with AI, he said.

“If we’re not having those discussions with our clients, then they’re obviously going someplace to learn that information because the adoption rate of ChatGPT was so drastic,” he said.

Timm pushed back against the notion that mass GenAI adoption is within three years for small businesses.

While that timeline makes sense for larger businesses with more datasets, small and midsize businesses (SMBs) will lean on MSPs for adoption.

“It is going to take us figuring out how to go to market with that to the SMB space,” he said.

Although the Meghan Landry chatbot is only used internally at General Informatics, Monisterem sees chatbots as a particularly helpful use case for AI within his health care clients, he told the crowd.

“It’s really difficult, especially in the health care space, for everyone to know exactly how they’re supposed to approach every business process that they have within their organization,” he said. “So it’s really handy for them to go and ask a chatbot, ‘Hey, how am I supposed to handle this specific situation?’ And if you train the model effectively enough – which is part of, really, the expertise that we bring to the table – you can get some really good information back to the end user.”