Why you want your services partner to embrace AI

with Greg Poirier

Compliance and legal teams can get nervous when they hear that a service provider is using AI, and that’s understandable, but Greg Poirier – SVP of Ateko’s Salesforce Practice, has some great insights on why you actually want your service provider to lean in on AI – and what benefits that can bring to your organization.

Today we’re here with Greg Poirier, SVP of our Salesforce Practice at Ateko and he’s got a hot take on AI in our services world. Greg, we know that some Service Providers and Implementation partners say they don’t use any AI, and that’s a significant issue. Can you explain why?

First off, I get why clients are asking – their legal and compliance teams are very worried about AI. And that makes sense. And I get why Service Implementers and Consultants are tempted to promise they won’t use any AI for an engagement. The issue is, both sides know that this is a false promise, everyone understands that AI is happening.

Anybody on either side of the engagement using Microsoft, Google, any office platform knows that AI is inherently baked in whether it be autocomplete in Word, or solutions like Cohere that are helping with coding. It’s happening.

So when we’re saying or agreeing that nobody’s going to use AI, we kind of know that we’re fundamentally lying about that.

Or alternatively, you’re working with a vendor who is operating their business vs. on an on-prem instance of Outlook and Windows 95.

More importantly though – AI makes projects go faster.vAnd if you are paying your vendor by the hour (which hopefully you’re not), but if you are… Do you want to pay them to convert a 100 page doc into a PowerPoint presentation for you by the hour? Or do you want them to leverage AI to do that?

Do you want them to leverage AI to make the one million record migration and transformation of data to your new Salesforce instance go much faster? And maybe get done in 2 weeks instead of 4 weeks? Or do you want to pay them by the hour, but also deliver your project much later.

So those are the inherent difficulties in stating that you don’t want any AI in a project.

Basically you’re saying you want everything to go slower, but also be more expensive. But what about clients who have some concerns due to the legality? Or their legal departments are saying that AI is just too risky to use?

Yeah, this comes up a lot with our clients, because primarily we service highly regulated industries, high compliance, high security clients.

So a lot of the time legal may say “hey we can’t use any AI” – And that’s ok, but usually we can have a conversation with them and say…

Well, are you ok with a world where someone starts typing an email and that sentence gets autocompleted? Probably.

Are you ok with a world where a developer begins typing a line of code and it’s a well-known one, it’s easily understood. That gets auto-completed. Ok. Maybe.

Are you ok with a world where AI is looking for common errors and reviewing an entire codeset or a project to detect where bugs may occur. That gets a little more complicated, but if we’re able to have that conversation and explain to them that this will allow the project to be delivered for let’s say 20% less the cost of a normal project, or their competitor’s, this will allow the project to be delivered 2 weeks faster, but perhaps most importantly… this will allow the project to be delivered with fewer bugs, fewer code exceptions, it runs more efficiently because these AI solutions can parse millions of lines of code that a human never could in order to find ways to make it more efficient or weed out possible errors and problems.

As you’re having that discussion, educating legal on the cost-benefit analysis of using AI is the important piece. And that’s where you start to get that buy-in.

So Greg, given your expertise in Salesforce and large enterprise adopting new technologies, what is the biggest  mistake that companies are making today?

The thing that we’re seeing most often is that enterprises are thinking about AI as “a singular platform” versus “this is now thing that is baked in to platform we use today”. And every platform our vendors use today.

So Greg, what is the best way that organizations are leveraging AI, in your opinion today?

I’m a big believer in the plant 1,000 seeds approach to AI. So I love to see when organizations secure a set of AI solutions that are safe for use, they issue guidelines and provide training to their employees, but then they let their teams figure out the best way to use AI to make them productive

Internally here at Ateko, we’ve got a mandate that we will be in aggregate 5% more efficient every quarter using AI So compounding – every project should take 22% – roughly – each time we use AI for it. We’re not going to achieve that by deploying one big AI automation project  (although we are doing those). We’re getting those gains by making safe AI tools available to every employee, and then celebrating the use cases they come up with in terms of using everyday AI in how they accomplish and how they do their jobs.

So what are you most excited about when it comes to the future of AI capabilities?

For me it is when I hear one of our team members say that: We’ve always done this long boring thing with X. Now, today – I’m using Cursor combined with Gemini, and I’ve shortened the amount of time that it took me to do this from one week to 4 hours. And so they’re able to take that boring work that is not very rewarding for them to do compress it, turn that into something that AI can easily accomplish and they’re coming up with very novel ways to do that, and then they’re able to use their time to do things that are more interesting and stimulating for them, and that’s a win-win. It’s a win for our clients because we can deliver projects faster. And it’s a win for our team members because they can focus on more interesting things to do.

Thanks Greg. That gives us a lot to think about.

We hope you found Greg’s insights helpful as you consider how your organization could – and should – be leveraging AI.