Amazon Quick Suite, Aussie SME Automation, and Deloitte’s $400k AI Slip

Llew Jury

Alright, welcome back to the AI Intelligence Podcast, with thanks to Advancer - The AI Agency. I'm your host Llew Jury and let's dive right in—big news from Amazon this week Ollie. They've launched what they're calling Quick Suite and it's pretty ambitious. They're promising full-on automation for enterprises by plugging into over 50 business apps right out of the box. We're talking heavy hitters: SharePoint, Salesforce, Google Drive—the whole lot.

Ollie Carter

Hi Llew and welcome everyone. Yeah, it is a bit of a flex, isn't it? I mean, their pitch is basically: let the AI handle your research, analysis, even workflow automations, so you don’t have to glue together all your business processes with duct tape and spreadsheets anymore. Quick Suite pulls from all your connected data and can run, like, real cross-system operations. And they’re saying it’s “agentic”—like an actual AI teammate for the business, not just a smarter chatbot.

Llew Jury

Spot on, Ollie. It’s not about flashy stuff like deepfakes or generating poems. Their goal is quite literally “AI that pays the bills,” yeah? Here’s a cracking example—they’ve got this customer, Propulse Lab, who slashed their customer service ticket handling time by 80%. That saved them 24,000 hours a year—just by plugging in those integrations. And then you've got DXC Technology, a global IT giant, rolling it out to more than 120,000 users. That's not just pilots or demos playing around—it's serious at-scale automation.

Ollie Carter

It’s a massive swing at the enterprise market. But here’s the bit I like: the platform isn’t locked up behind technical wizardry. The Model Context Protocol or MCP as it's called, lets them reach out to hundreds more apps—so, it’s less about being a data scientist and more about getting work done. Even QuickSight for BI, QuickResearch for, like, automated analyst work, QuickFlows for automating those gnarly repetitive tasks. It’s all pretty slick—on paper, at least.

Llew Jury

Exactly. And it’s not just a US thing—this lands for Aussie businesses too. There are plenty of Australian businesses that I'm seeing with half a dozen different business systems that never speak to each other. You can now go in an set up automation pretty similar to Quick Suite, they saved weeks of reporting time each month. Actual, measurable time back—stuff that helps the bottom line. Where was I going with this? Oh, right... This is the kind of thing we've wanted to see: AI that's good for business, not just tech headlines.

Ollie Carter

I keep thinking back to what we said in Episode 10: "real AI wins are the ones that solve pain points, not just add digital noise". Quick Suite’s a great example of aiming for that real-world impact. But yeah, gotta ask: how much of this is deliverable hype, and how much is genuinely transformative once you’re past the pilot phase?

Ollie Carter

Alright, before we go full tinfoil hat on enterprise AI, let's ground it with some Aussie SME wins—because, honestly, that's where the magic's happening, right? I was looking into IOTAI’s recent projects—these guys specialise in automation for smaller businesses. Love what they did for a finance firm in Adelaide: automated their client onboarding, saved 40 hours a week. Full-time employee freed up, and it paid for itself in four weeks. That’s a 220% ROI, not just some theoretical “it’ll pay off by 2030” thing.

Llew Jury

Yeah, mate, and don't forget Melbourne’s Bella Vista Fashion. They were completely drowning in inventory chaos. AI stepped in, sorted it all out, and—bam—25 hours a week saved, 142% ROI. Owner reckons it changed the way the whole business ran. And the best bit? They didn’t need some NASA-level tech stack to make it work.

Ollie Carter

Right, right, all this was done with off-the-shelf, low-code platforms—Retool, n8n, stuff you can put together in a few weekends if you know where to look. I've heard of startups doing something similar recently —they were getting buried by basic customer service queries. They set up a simple AI-driven chat workflow, and suddenly their lean team had room to actually build new stuff, not just fight fires all day. It sounds small, but these “easy wins” seriously add up. Bang for buck, it’s hard to beat.

Llew Jury

Yeah, mate—big lessons here for anyone listening. You don’t need a giant budget or an army of engineers to start automating. The tools have never been more accessible. Figure out which repetitive task is burning you out, automate that, and you’ll see results pretty quick. It’s the real “AI pays the bills” approach—basically what we keep banging on about, episode after episode.

Ollie Carter

Exactly, Llew. And it actually backs up what we found in Episode 8 as well—start small, keep it focused, don’t get sucked into all the hype, and you’ll see proper ROI. Everyone loves a moonshot story, but incremental automation is where the actual business wins stack up.

Llew Jury

Alright, now for the facepalm portion of today’s show—AI gone wrong, or as folks love to call it now, “workslop.” Deloitte’s $400,000 AI-generated report for the government—have you seen this, Ollie? It’s wild. Fake academic references, non-existent quotes, and even a made-up book credited to a real professor. We're talking the kind of stuff you’d fail a uni assignment for, right?

Ollie Carter

It’s actually gobsmacking, Llew. Deloitte’s not exactly a backyard operation, and they let AI hallucinate its way through a government project! Harvard reckons 40% of employees are now running into this sort of dodgy AI work—looks fine on the surface, but underneath it’s, like, totally flawed.

Llew Jury

Agree Ollie, it doesn't matter if it's a $400 or $400k project, you need a human in the loop, or it’ll bite you, guaranteed.

Ollie Carter

Yeah, not to harp on it, but it’s not just about checking for typos. If the AI can make up quotes and sources for Deloitte, what’s stopping it in your everyday ops? The reputational fallout is so much bigger than just a refund. Like, it can torch trust your business spent a decade building basically overnight.

Llew Jury

Let this be the big red flag for everyone listening: AI can supercharge you, but it can just as easily send you off a cliff if you’re not paying attention. Human oversight isn’t just nice to have, it’s non-negotiable, especially when it comes to anything external or regulatory-facing.

Ollie Carter

Couldn’t agree more, Llew. Alright, that's probably a good place to wrap for this week. The main thread tying this all together? Practicality. Whether you're thinking big—like Amazon’s Quick Suite—or you’re just trying to solve day-to-day pain with off-the-shelf tools, keep the human element front and centre. Test, review, and remember AI won’t fix bad business processes. Anything else before we sign off?

Llew Jury

That’s pretty much it from me! Thanks for tuning in, everyone—Ollie, always good yarn. We’ll be back next week to break down more AI hits and misses.

Ollie Carter

Cheers, Llew. Catch you all next time. And don’t forget to check out the team at Advancer if you want help on your own AI journey. See ya!