This episode starts with a familiar scene. A role opens, the applications pour in, and suddenly you're staring at a mountain of resumes that deserve real attention but arrive faster than anyone can process. The mix had everything… experienced candidates, newcomers trying to break in, and a growing stack of AI-generated submissions that looked sharp until you asked a second question. That's where Haystack came in. Instead of using AI as a blunt filter, Rob and the team treated it like a collaborator. Teach it what matters. Teach it what P3 looks for in a teammate. Teach it how to separate real signal from polished noise. What came back wasn't a robot recruiter. It was clarity. And Haystack is only half the story. As the conversation unfolds, Rob and Justin zoom out into the broader pattern they're seeing across all the small, useful agents taking shape inside P3. The stuff that isn't blind hype. The stuff that quietly fixes overloaded parts of the business and makes the human decisions easier to get right. Because that's the through-line here. When AI handles the overflow, people get to spend their time on the work that actually requires judgment. Queue it up and hear what happens when AI stops pretending to be magic and starts doing real work. And if you've got a corner of the business that's begging for that kind of clarity, we can help you find the tiny build that changes everything.
Rob finally cracked his years long standoff with the podcast lair cat, and the fix was hilariously simple. That small victory ends up setting the tone for the whole episode, because everything that follows has the same energy: real problems that only make sense once you shrink the solution down. As Rob talks through the cat truce, Justin brings in a different kind of chaos. A customer service bot that sounded fully in command yet never actually did the thing it said it did. Pair that with a hiring queue full of AI written applications, and the whole picture starts to come into focus. Once you see the pattern, you can't unsee it. The wins only show up when the AI job gets small. The fantasy football tool works the moment AI stops trying to scrape the entire internet and instead only writes the human part. The hiring filter works when AI is there to catch repetitive patterns, not run the whole show. Even the experiments coming out of Danielson Labs click only because the AI calls are tiny and the real work sits in regular code. Everything points in the same direction. Let AI handle the one thing only AI can do, then let normal tech take it from there.
Everyone's talking about AI like it's plug-and-play. Spoiler: it's not. In this episode, Rob digs into why Big Tech's billions in AI R&D haven't yet turned into matching revenue — and what that means for the rest of us. The truth? The real business wins don't come from off-the-shelf models; they come from smart customization. Rob breaks down the "magic Lego brick" approach that separates hype from practical reality, showing how everyday tools like Power BI and Power Automate can connect to AI in surprisingly simple (and powerful) ways. He also revisits Bill Krolicki's "Vendor Bot" example to prove that you don't need to be a researcher or a billionaire to make AI deliver real results. If you've ever opened ChatGPT, asked it to "optimize operations," and gotten nowhere — this one's for you.
When the bubble pops, what's left standing? Everyone's calling AI a bubble — and yeah, there's some wild stuff happening behind the scenes. Billions changing hands between the same few tech giants. Investments that look more like circular loans than innovation. But what if the real bubble isn't AI at all? Rob breaks down what's really inflating, what's not, and why your business shouldn't lose sleep over it. The headlines may sound like the sky's falling, but the tools driving real results aren't going anywhere. Some bubbles burst. Others just float higher. Listen now to hear Rob's take on which kind this is and why the real value of AI doesn't care what the stock chart says.
Everyone's talking about AI. Almost no one's getting specific. And that's the problem. Rob and Justin break down the three categories that matter: chat agents you talk to, embedded assistants baked into your workflows, and headless systems running quietly in the background doing the grunt work nobody wants to do. This isn't theory. This is how AI is finally working for real companies. They walk through Vendor Bot, Scheduler Bot, and Budget Bot — small, focused tools that do one thing exceptionally well. Because here's what keeps happening: the "one bot to rule them all" projects collapse under their own weight every single time. But stack a few narrow, reliable bots together? That's when things get interesting. If you've been wondering why AI works for some teams and stalls for others, this episode has your answer. Spoiler: it's not about going bigger. It's about getting ruthlessly specific. The "we're all in on AI" era is over. The "we know where it fits" era just began. If you like the show, be sure to leave us a review or share it on your platform of choice.
AI without the committee: Most manufacturers are still holding meetings about AI. Bill Krolicki just built it. As CFO of Interpak, he didn't wait for a strategy deck or a vendor pilot—he wired Power BI and Power Automate into a self-running operation. His bots read supplier emails, catch late shipments before they blow up production, and update the ERP while everyone else is still asking who owns the spreadsheet. Rob Collie and Justin Mannhardt sit down with Bill to talk about what happens when finance stops waiting and starts building. From "Vendor Bot" to the soon-to-launch "Budget Bot," it's a front-row look at how AI turns from theory to throughput when a data person is actually in charge—no consultants, no six-month roadmaps, just results. If you've had your fill of AI hype and want to see what it looks like when someone actually ships something, this is your episode. And if you enjoyed it, leave us a review on your favorite podcast platform—it helps other no-BS practitioners find us.
Last week we got a facelift—new name, new look, same deep data dives. This week? We prove the rebrand wasn't just cosmetic. Rob kicks things off with a time machine moment: his first gig at Microsoft in the '90s, building the Windows Installer. The running joke back then? "Installing yesterday's apps tomorrow." Cut to 2025, and that exact same code shows up while he's configuring an AI tool for data modeling. Build something right, and it really sticks around. And that's the bridge, AI context management isn't some brave new world. It's the same discipline that made Power BI models and Copilot integrations actually useful. You don't need to burn it all down and start over. You just need to get specific enough to matter. If you've suffered through bloated "AI strategy" decks or watched a model confidently hallucinate through your business logic, this episode's for you. The fix isn't fancier AI—it's giving it structure, purpose, and the right context to work with. That's how you turn a show pony into a workhorse. Bottom line: AI isn't a revolution. It's a new faucet. And the people who know how to connect it—and what to feed it—are already leading the next wave of transformation.
Everyone's talking like AI is coming for your job. Rob's here to tell you it's coming for your skill set. If you already think in tables, models, and relationships, congratulations—you're built for what's next. This episode is part pep talk, part reality check, and all proof that data people aren't getting replaced. We're getting promoted. Rob takes you inside Microsoft's campus and out into the real world, where big firms are burning millions on AI theater while mid-market teams quietly pull ahead. The twist? You don't need to change the AI to win with it. You just need to feed it the right data. That means the same instincts that made you good at Power BI now make you dangerous—in the best way. This one's for the data geners, the ones who never flinched at a gnarly spreadsheet and always saw potential where others saw pain. AI isn't a threat. It's your next playground. Listen in and meet the only cavalry that's actually showing up: you. Check out the first episode of Raw Data with Rob Collie and get ready to lead the AI charge.
Nobody loves requirements docs. They're the corporate equivalent of writing a novel just so someone can skim the back cover. The real question is whether you can ditch all that and go straight from "here's what I need" to a working Power BI model. In this episode, Rob and Justin push AI into that role and see what breaks, what builds, and what actually saves you time. Turns out, the magic isn't in making AI look impressive on a demo slide. It's in whether it can wire up tables, relationships, and measures fast enough that your team can skip the plumbing and jump right to the good part: asking "does this answer the question?" instead of "why won't this table join?" That's the test, and it's the only one that matters. From tools that feel like friendly appliances to those that lean full hacker-mode, Rob and Justin run the gauntlet. They even crack open Copilot's inner workings to see how answers really get formed. It's a gritty look at whether AI can finally cut the "first-hour tax" every project pays and give leaders a faster path to value.
Large language models aren't magic. They're pivot tables for words. That's the real breakthrough — not a crystal ball, not a robot overlord, but a new way to roll up all the noise in your business into something you can actually use. And that's why AI belongs in the middle layer. Just like BI gave you visibility across systems, AI is becoming the connective tissue for all the unstructured stuff that never fit neatly in a database. Sure, every product is rushing to bolt on "AI features," but those sidebars and pop-ups can only see the data inside their own walls. The real power shows up when you wire AI across the mess — the emails, the docs, the meeting notes, the structured and unstructured side by side. That's where pivot-table-for-words meets pivot-table-for-numbers, and suddenly you're not staring at silos. You're staring at the whole picture. Rob and Justin cut past the hype to show why AI isn't the star of the show, it's the glue in the middle. And that's good news, because the middle is where business actually gets done. Hit play and hear why the future of AI is less wizardry, more wiring — and why that's exactly what makes it work.
This isn't another AI think-piece, it's a full-on data brawl. Copilot is out here plagiarizing Rob's pivot table crusade while the self-appointed nerd police try to lock down the definition of agentic AI. Meanwhile, thirty years of fantasy football become the unexpected proof that tuning beats buzzwords every single time. What starts as a slip from Copilot turns into a bigger story about how AI really works. Off-the-shelf tools can sound impressive, but they collapse into clichés when they're not tuned to the person using them. The difference isn't just in efficiency, it's in credibility. Get it right, and AI amplifies your voice. Get it wrong, and you sound like everyone else mailing it in. Don't settle for AI that sounds like everyone else. Listen in and hear what happens when tuned workflows collide with real-world stakes.
Pivot tables finally auto-refresh. Humanity wins, but our host Rob Collie is still annoyed. Why? Because he asked for it back in 2007 and got shot down. That little gripe kicks off a bigger conversation: what else in data is overdue for a shake-up? Copilot might be next. Forget filters. Forget dashboards. Rob put it to the test on his beer-league hockey stats and found himself asking questions faster than any report could keep up. It felt natural, almost too natural. Then came the twist. Copilot served up an answer that looked perfect… and was completely wrong. If it can fool the guy who built the model, what chance does anyone else have? Listen in for the laughs, the lessons, and the curveballs of a future where filters may finally be obsolete.
Every campaign has it: that shiny "more reach, lower cost" lever that looks like marketing gold but really just siphons your budget into digital quicksand. We pulled it. The metrics looked fantastic. And that's exactly how we knew we'd been had. In this episode, Rob breaks down a real-world lesson in false signals, phantom clicks, and why data discipline isn't just consultant-speak—it's your financial survival strategy. We're talking about the bots that sneak in through "partner sites," how they corrupt your retargeting, and the ripple effect that turns good data bad across your entire funnel. Here's the thing: The real damage isn't the wasted spend. It's what corrupted data does to your decision-making. When you're building your strategy on phantom signals, every "optimization" takes you further from real results. If you're protecting a budget or leading a team that depends on clean data to drive real business decisions, this episode cuts straight to what matters: spotting the traps before they drain your resources and rebuilding trust in the numbers that actually count.
AI is rewriting the rules of analytics. Copilot can pull answers straight from your semantic model and bypass the dashboard entirely. But for all the tech fireworks, the same old truth holds: communication is still the hardest part. Stakeholders don't always know what they want, builders don't always know how to translate it, and requirements docs have never fixed that gap. Copilot just puts the tension in sharper focus. Rob and Justin dig into why vanishing chat histories aren't just inconvenient, they erase the most honest record of what stakeholders actually care about. Screenshots and Word docs are a band-aid, not a solution. Persistent, shareable conversations could change the way model developers and business users collaborate, but only if governance and security evolve fast enough to keep up. Along the way, they show why usage data from Copilot queries is miles ahead of click stats on a dashboard and why the story of your data has always hinged on the same thing: people understanding each other. Dashboards may have set the stage, but conversation is where the real action is. Listen now and see what happens when the chat itself becomes the deliverable.
Two hundred episodes in, and we're done with the warm-up! Episode 200 finds Rob flying solo and pulling zero punches on the question everyone's quietly asking: what's the future of data work? No anniversary nostalgia here, just uncomfortable truths about AI bootcamps at Starbucks, semantic models going naked, and why being "pretty good" at anything is about to get very complicated. If you think you know where this is all headed, think again. Rob spent yesterday turning a non-techie real estate agent into an AI power user, and what he saw exceeded anything from the early Power BI days. But here's what nobody's talking about: the invisible barriers, the shifting skill requirements, and why the middle ground might be disappearing faster than anyone realizes. The lines between data work and software work are blurring, structured versus unstructured data is becoming meaningless, and the comfortable assumptions about who does what are about to get stress-tested. Two hundred episodes of calling it straight, and this one tackles the questions that keep data professionals up at night. Some answers might surprise you. Others might make you uncomfortable. [But you'll know exactly where you stand when the dust settles]
Most of us have been in the trenches long enough to know when something's about to flip the script. And brother, we're standing at the edge of a cliff most data folks don't even see coming. Rob Collie thought he had Power BI figured out. Then Copilot did something impossible; it cracked a question that should've left it scratching its digital head. But it didn't just answer. It nailed it. That's what we're calling the Dobie Moment—when AI stops being a fancy calculator and starts being genuinely scary-smart. Here's the thing nobody's talking about, your semantic models aren't just sitting there anymore. They're waking up. And when Rob and Justin break down what happened in this episode, you'll see exactly why that should make you sweat a little. They're not here to blow smoke. They'll show you the magic, sure, but more importantly, they'll show you where the landmines are buried. Because when AI starts connecting dots you didn't even know existed, confidence and correctness become two very different animals. Bottom line: The future of data just knocked on your door. You can pretend you didn't hear it, or you can listen to this episode and actually be ready when your models have their own moment of reckoning. Your call. But don't say we didn't warn you.
AI looks unstoppable… until you hand it a hundred pages of meeting notes. Rob and Justin dig into why context windows and token limits quietly run the show. That "million-token" brag from Google? More like weighing the Titanic in bananas. From Shakespeare to SharePoint, this episode shows why AI remembers the Roman Empire better than your company history—and why that's not a bad thing. Rob also introduces Griff, a digital colleague that fires off P3-flavored ideas like it's had three espressos. It's practical AI that's actually fun to use. Hit play to find out where AI is brilliant, where it falls flat, and how to make it work for you without the hype. Also on this episode: Million Token Context Windows? Myth Busted—Limits & Fixes
Back in 2010, Tableau beat smarter tools with a better demo. No brain, all charm and the market loved it. Fast-forward to now: same playbook, new costume. The AI dashboard crowd is selling "natural language BI" with zero semantic model, zero memory, and a whole lot of LinkedIn swagger. In this episode, Rob and Justin revisit why Tableau's empty-calorie approach won the first round, and how that same mistake is about to flood the AI + BI space all over again. Turns out, you can still sell snake oil if you call it GenAI. Rob breaks down how an elite MIT course managed to skip LLMs entirely, how a flashy Tableau blog post went viral for connecting a CSV, and why "AI-ready" vendors keep duct-taping chat interfaces onto raw SQL and hoping no one looks under the hood. But the real story? Microsoft is sitting on the most powerful data brain in the game, and if they land the front end, it's game over. This isn't just a history lesson. It's a blueprint for seeing through the hype and betting on what actually works. If you're building, buying, or betting on AI tools, listen in before you get dazzled by the demo. Also on this episode: Early Experiments in Tableau's New MCP Service
Let's say your business runs on sun, sweat, and schedule precision. You've got crews in the field, materials that don't wait, and about 90 minutes to get it right before the product turns into a thousand-pound paperweight. That's the world Joseph Graziano lives in. He's the CFO of Quadrant Concrete and also the guy keeping the trucks moving, the forecasts dialed in, and the safety records spotless. Because in the mid-market, you don't get extra people. You get extra resourceful. Joseph shares how he helped transform a boots-on-the-ground concrete business into a data-forward operation without fancy titles, inflated budgets, or a fleet of consultants. From field-collected data to real-time cashflow forecasting, he's found the sweet spot where better reporting leads to smarter, calmer decisions. You'll also hear why operational transparency isn't just about ROI, it's about reducing chaos, building trust, and creating a culture where everyone sees what matters. If you're trying to lead your business through complexity without adding complexity, this episode can be your blueprint. Data doesn't have to be fancy to be powerful. It just has to work. Listen now and see what it looks like when grit meets insight.
It started as a side project. Rob Collie built a Power BI model for his rec league hockey team. Just for fun. Just to see what the data could say. But something weird happened. The dashboards were solid. The data model was solid. But, the users still had questions. And lots of them. And that's when it clicked: people don't think in slicers. They think in questions. Natural ones. The kind dashboards rarely anticipate. In this episode, Rob and Justin Mannhardt didn't just talk about Microsoft's Copilot for Power BI. They put it to the test. No tuning. No prep. Just a raw semantic model paired with real questions from actual humans. The result? A glimpse at what happens when the tech finally meets the moment. Copilot isn't just a gimmick. It understands nuance, handles filters, and points people to the answer without making them dig. And it's getting better by the day. This isn't a future-state conversation. You've already done the hard part. Now you can build on it. And if you've been wondering when AI will start delivering real value, this is a pretty good place to start. Also in this episode: Indy Inline Hockey Dashboards Inline Analytics Doesn't Mean What You Suspect it Means, w/Ryan Spahr Copilot for Power BI Rethinking the ROI of Dashboards