Monte Carlo

Are you busy on November 6th? You are now! IMPACT: The Data + AI Observability Virtual Summit is back on November 6th and trust us, you won’t want to miss it. Now’s your chance to learn from top data + AI leaders from across industries to explore how they’re scaling trust, reliability, and observability across their modern data + AI stacks. Join us to learn how data + AI teams at M&T Bank, Warner Bros Discovery, T. Rowe Price, Pilot Flying J, and Roche are: -Bridging the gap between data and AI in production -Ensuring visibility and accountability in increasingly complex architectures -How observability is enabling trusted, high-impact AI initiatives And so much more Spots will fill up, so save yours now! RSVP here: https://hubs.li/Q03C0zGg0

Are we living on a technological fault-line? Almost as soon as the announcement came down about OpenAI’s AgentKit (old news by now), the LinkedIn rumor mill was churning out the epitaphs of whole swathes of new and established tooling. Transformative one day. Archaic the next. But here's the big question: is the world really so different? Or have the standards for creating value just resettled within their natural boundaries? Here's the situation: if AI makes it possible for anyone to vibe-code anything, then what we choose to build or buy will need to be commensurately more discriminating. And I know, AI code is buggy and most of it never makes it to production, but just follow me for a second. Even if the calculus of how a product gets developed changes, the common denominator for what makes a product valuable won't. In the land of the blind, the one-eyed man is king. But in a world dominated by AI, experience still wins the day. Here’s my hot take: if you're using AI to solve the problems that someone else can solve, then eventually they will. So, use AI to solve the problems that only you can solve. In other words, if you want to deliver impact, focus your energy where you can establish unique value based on what’s proprietary to your organization—which more often than not, comes down to the context data it owns and the teams that support it. Choosing the right AI pilot isn’t about creating the flashiest demo—it’s about bringing your own unique assets to bear on a technology that can optimize them. And the same goes for your partners. In an AI environment, iteration cycles happen quickly. But regardless of how fast you can build, what you can’t acquire quickly is context. What you can’t acquire quickly is an experienced perspective on the problem. Vibe-coding can certainly help you build faster, but faster doesn’t mean more valuable. If you’re looking for a partner that’s still going to exist in 5 years, you need to look for companies that check those boxes first. AI doesn't change value—it clarifies it. Whether you’re deploying pilots internally or evaluating partners externally, reliable proprietary context data—and the expertise to know how to use it—is the first and only moat for value in the AI era. More on this topic in my latest article (check the comments!)

Check out our very own Devon Bridges, SDR Manager at Monte Carlo, featured in Built In's round up of sales leaders fostering engagement on their teams! "Collaboration is critical; no one wins alone, so I create a culture of shared learning, call blitzes and cross-functional alignment. Curiosity is a superpower, so I push my team to challenge assumptions, share insights and keep learning. Above all, I lead with intentional energy and positivity. Engagement happens when reps know I have their back and future in mind, and that’s what earns me the right to push them to be their best." - Devon Bridges Check out the full article here: https://lnkd.in/eJsUjTb8 We're hiring! Check out our open roles: https://lnkd.in/eXqfg2tm #sales #salesleadership #SDR #BDR #techsales #data #AI

How good is “good enough” for production? Interesting interview CDO Magazine series with Wendy Batchelder from Centene Corporation about how her team is approaching AI production that hits on so many of the issues data and AI teams are facing in the move to production: - Trust - Transparency - Governance - Fragmentation - Quality - Speed - Ethical standards And in the context of AI-ready data, “perfect” isn’t a viable standard for any one of these... not just because it’s prohibitively expensive—but also because it's functionally impossible. So, what IS acceptable? When is the data ready? When is your team ready? I think this response from Wendy summarizes the answer pretty well: "Prioritize: where are we going to be using data? Which use cases are most important? What level of quality is acceptable, and do we have end-to-end transparency to really understand — where did this data come from? How has it been modified, manipulated, changed, or curated?” We live in a world of opportunity costs. We can’t do everything. But in an AI arms race where getting to value faster can mean the difference between legacy and irrelevance, the best thing we can do is ruthlessly prioritize for impact. That requires two things: - Getting closer to your business - And getting closer to your data.  That means establishing technologies and processes that make it easier to quantify who’s using the data, what they’re using it for, how often they’re using it, and how performant it is for those use cases at any given time. Tactics like observability, data product thinking, catalogs, and SLAs can all be important here. But it starts with the business. Bottom-line is, if you can’t align standards to use cases and use cases to value, the only thing vibe-coding is going to give you is AI pilots that fail even faster in production. Link to full article in the comments!

"We've built incredible data systems. But we've forgotten how to trust them." As data products become embedded in core business operations, a dangerous assumption has taken hold: if the pipeline runs without errors, the data must be correct. It's not. I see this pattern repeatedly with clients, green checkmarks across dashboards while millions flow through systems built on subtly corrupted data. A successful run tells you the code executed. It doesn't tell you whether last Tuesday's revenue spike was real or a duplicate load. The problem is about to get worse. As we layer AI onto these systems, the distance between "it ran" and "it's right" will only widen. Graph - Michael Segner, 2023

The latest edition of The Observer is here, just in time for spooky season. Catch up on all the latest happenings in data and AI... if you dare... 👻 🎃 And don't forget to subscribe to get these updates straight to your inbox every month! #data #AI #agents #observability #dataplatforms #dataproducts

We'll be at the Databricks World Tour in London tomorrow! Visit us at Booth #5 and be sure to stop by Karthick Sriraman's session at 15:00 BST. He'll explore why 2026 is the year of data + AI observability and highlight key trends driving this transformation. Learn how observability bridges the gap between data and AI systems across your data, system, code, and models, enabling more trustworthy, scalable, and efficient operations. Learn more here: https://lnkd.in/dEWZw4Ea #DBX #Databricks #WorldTour #observability #dataquality

Thank you to everyone who joined the Gartner C-level Communities Boston CDAO Community Town Hall yesterday! It was incredible to see 50 of Boston’s brightest minds come together to discuss the topic “Accelerate AI Readiness: Scaling Quality With Data + AI Observability.” Some key takeaways from our discussion: ➡️ Choose use cases that matter—focus on initiatives that drive real business buy-in. ➡️ Don’t die by a thousand AI cuts—with so many vendors, it’s easy to try to boil the ocean. Stay focused! ➡️ Data stewards matter—having the right data in place is critical. I'm grateful for the engagement, insights, and energy everyone brought to the conversation. Extra special thanks to Monte Carlo for sponsoring this incredible gathering. I would also like to thank Barr Moses, Karthik Yajurvedi, Manajit Barman, Naveed Afzal, Ph.D., and Manish Nigam for leading the insightful discussions. Looking forward to continuing this journey together! #AI #DataObservability #CDAO #GartnerClevelCommunities #BostonDataLeaders

Bad data is killing your agents. Or is it the model? How would you know? Data quality only validates the data before it's embedded. Evaluations only cover the last mile of your agents. If you can't observe the entire pipeline from data to output, you can't make that agent reliable in production. AI observability may start at the model—but it can't end there. The ability to understand what happened, where in the pipeline it happened, and how to manage it at scale is the essence of observability. It's also the only way to deliver reliable agents in production. And the truth is, most of what masquerades as hallucination is really just a data issue in disguise. Learn more about my POV here: https://lnkd.in/gT_kGdpY

Don’t forget: you’re busy on November 6th. 🚀 Join us for IMPACT 2025: The Data + AI Observability Summit, where we’ll dive deep into what will define the next era of trusted data + AI. Whether you’re leading enterprise AI initiatives, managing large-scale data platforms, or tackling governance and compliance, this is a can’t-miss event to stay a step ahead. Data + AI leaders from Roche, Pilot Flying J, T. Rowe Price, M&T Bank, and more will be sharing: ✅ Tools for building resilient data foundations ✅ Frameworks for bridging the data + AI reliability gap before and during production ✅ Real-world strategies for operationalizing observability across data and AI ✅ Lessons from top-performing organizations scaling AI ✅ A first look at the latest innovations pioneering future of data + AI observability ✅ And more! Don’t miss your chance to lead the reliable data + AI charge at your organization. RSVP here: https://lnkd.in/eu6vsCfn #IMPACT #observability #dataquality #agents #AIquality