alunablue - stock.adobe.com
The AI journey must start with the data
MSPs advised to begin conversations with customers at a data level to ensure AI tools will deliver the promised results
For the channel to help unlock the potential of artificial intelligence (AI) for customers, the starting place has to be data, and ensuring customers have a solid base from which to draw insights.
The old adage “crap in crap out” is one danger with AI systems, as is the prospect of it being limited to a handful of executives who want to unlock the promise of the technology without taking advantage of the insights that could emerge from across the wider organisation.
Those themes were discussed earlier this week at an MSP event held by AvePoint, which is keen to underline that getting a data strategy established is the best place for partners to start with AI journeys.
Chris Shaw, UK channel manager at AvePoint, said that use cases for AI are emerging, Microsoft is driving the market forward with Copilot and users are hungry for the technology. But there is a danger that just dropping in tools and waiting for inspirational results will put the MSP in a position where the odds are stacked against it.
“If you look at it from an end user’s perspective, there are four or five stages you can take us to,” he said. “The first one is getting the data into the right place, whether that’s coming out of legacy systems or whether that is coming from multiple platforms.
“You migrate the data, and from there, you can then apply security, governance and control, and that’s where our AI data security piece starts to fall in and become very valuable.
“Everything that AI does is actually based on data,” said Shaw. “So, if your data is not controlled, if your data is not secure, if your data is not in a usable format, it actually gives you really poor results. And that’s actually what I think is holding a lot of people back.”
Talking about data
Even those that are not finding customers are dragging their heels on AI need to talk about data because of the long term benefits of gaining a grip on their information.
“Even if you haven’t got Copilot, or if you’re not talking about AI, that data privacy, that modelling of getting your data in a good footprint is actually good practice anyway,” he said. “Do you have the right level of security tests? Do you have the right permissions for sharing all of those sorts of things, and those topics are really, really relevant today – even if you don’t have copilot.”
Shaw said there was a lot of focus on AI but the channel still needs to play an educational role guiding customers through their first steps with the technology.
“People don’t necessarily know where to start,” he said. “It’s a big ask, and I think as a conceptual problem, if you say to an end user, ‘Copilot will fix your problems,’ it’s quite hard to understand what problems you’ve got to know to fix.
“I think people are only just now starting to define that problem in their minds and then the associated actions of the back of that AI data readiness,” said Shaw.
