5 takeaways from Precisely Now: Trusted AI starts with trusted data

Key Takeaways 

• AI moved from planning to production faster than data readiness did, and agents now act on untrusted data at machine speed, before anyone can intervene. 

• Precisely announced the Precisely Platform for unified data management and Precisely AI Studio for building on it. AI Studio is available today. 

• Customers at Lake Michigan Credit Union (LMCU) and Charter say the hard part is shared definitions and governance, and they advise starting with a business problem, not with AI.

Eighty-six percent of organizations have increased their focus on data management because of AI. Yet most AI initiatives are still struggling to reach production. 

That tension ran through every conversation at Precisely Now: Where Data Meets AI. I hosted the event, and between our CEO’s candid conversation with IDC about what buyers are getting right and wrong, the product announcements, and an engaging customer panel, a clear picture emerged of what AI-ready data requires.  

These are the five takeaways that stuck with me most.

1. AI moved into production faster than trusted data could follow

When our Customer Advisory Board met last November, the mood was cautious. These are twenty of our largest customers across industries including financial services, insurance, and telco. And the message was plain: we’re not ready. Six months later, those same customers had initiatives in flight, and the question had shifted from “should we use AI?” to “why isn’t this working yet?” 

The answer almost always came back to data. IDC’s research points the same way: in a survey of more than 1,000 respondents, 44% of AI proofs of concept that never reached production cited data quality as the primary reason, and another 33% said it played a significant role. That is over three-quarters of failed projects traced back to data. 

AI projects stall in production mainly because the data underneath them isn’t ready, not because the models are wrong.

2. Agents are the new data consumers, and they work at machine speed

Our CEO, Walid Abu-Hadba, has led through more than one technology shift throughout his career at companies from Microsoft to Oracle to Sage and now Precisely, and he explained why this shift is different from the ones he’s seen.  

Walid pointed out that technology used to be the hardest part to get right, but with AI, you also need trusted data for that technology to operate on. The issue is that most enterprise data is scattered across mainframe, on-premises, cloud, and SaaS systems, some of it stale or inaccurate. Unleash AI on that data and you won’t get a trusted outcome. As he put it, “If you want trusted AI, you need trusted data.” 

He then turned to Stewart Bond, author of the new 2026 IDC MarketScape for Worldwide Data Intelligence Platform Software, for an honest read on what’s changing, what buyers are getting right and wrong, and where the real opportunity is. 

Stewart pointed out that data used to feed dashboards and spreadsheets that people read. Now it feeds agents making multi-step, autonomous decisions, and they can’t wait for data to be ready. It has to be ready when the machine needs it. 

That changes the job of data intelligence. It once told you where your data was and whether it was clean. Now it has to answer whether an autonomous agent can trust that data enough to act without a human checking first. Stewart’s test for governance leaders is simple: can I see what it did, and can I undo it? 

When agents consume data at machine speed, trust has to be established before the agent acts, with explainability and human oversight built in. 

3. Unified data management replaces the point-solution stack

For years, organizations ran separate tools for data quality, governance, cataloging, and enrichment, with analytics stitching them together. That worked until AI. Add AI at every one of those points and a complicated setup becomes chaotic: agents pulling from fragmented data in different states of quality, with no shared governance underneath. 

That’s why we were thrilled to share more about our newly announced Precisely Platform, now available in preview. The platform brings together data integration, data quality, governance, location intelligence, master data management, and customer communications into one experience. Three design choices stand out. 

• It’s a control plane rather than a data plane, so it reaches data in the cloud, in lakehouses, and on-premises without a migration. We built it this way because the systems that run your business can’t always move, and AI needs to reach them where they are. 

• A unified semantic layer gives agents the context to make more deterministic decisions.  

• And 9,000 data attributes across hundreds of proprietary datasets sit inside your quality pipelines, so enrichment isn’t another tool to buy and integrate.  

Walid underscored why this choice matters. As he put it, “It’s so hard to go back if you’ve picked the wrong architecture.” Stewart’s advice was to start from requirements: if your data spans clouds, legacy systems, and applications, hybrid execution can matter more than any single ecosystem’s native depth. IDC named Precisely a Leader among the 15 vendors it evaluated. 

Unified data management gives AI applications, agents, and analytics one trusted foundation to operate from. 

4. Precisely AI Studio gives teams a place to start building faster

A platform is only as useful as what you can build on it. Precisely AI Studio, also just announced, is the entry point for the Precisely Platform. It’s where data teams, AI engineers, and developers access ready-made apps, agents, and skills to discover what’s possible and accelerate their own building.  

Building a demo is easy. Building something you can trust at scale, inside a complex enterprise, is the hard part, and AI Studio lowers the barrier to getting started. Customers and partners will contribute their own over time, making AI Studio a community of builders sharing tools with one another. 

It’s also MCP-native, so it connects to the model providers, tools, and applications your teams already use. It works with the existing Data Integrity Suite and the new Precisely Platform, which is what makes whatever you build trustworthy. 

We also gave an early look at a multi-domain master data management (MDM) workflow built on the Platform, and we’ll work with a select group of design partners on it this fall. 

AI Studio is where teams explore the possibilities and accelerate AI development. And it’s all grounded in data that the Platform has already ensured is governed, enriched, and trusted. 

5. The hardest work is shared definitions, not technology

Stewart noted that the root cause of data problems is often people and processes. Our customer panel, moderated by David Woods, our SVP of Global Services, made that concrete. 

Kevin Guyot of Lake Michigan Credit Union (LMCU) said access to technology was never the obstacle. The obstacle was discovering that teams defined “customer,” “product,” and “member” differently without realizing it. LMCU built a shared governance model and trained front-line teams in data literacy. What stood out to me is that his biggest hurdle wasn’t technical at all. 

Jeannine Roberts of Charter Communications made the case for context: reports read without it lead to wrong decisions, at AI speed and scale. Her approach to a strong data foundation for AI? “Test it, govern it, retest it” in small segments that show value.  

Kevin’s advice for data leaders: start with real business challenges rather than starting with AI. Successful AI programs begin with shared definitions, governance education, and a specific business problem. 

What should data, analytics, and AI leaders take away from Precisely Now? 

For years, organizations could work around fragmented data. People filled in the gaps, teams reconciled inconsistencies, and subject matter experts supplied the context systems lacked. AI systems don’t do that. They act on the data they’re given, at scale, faster than anyone can step in. 

That’s the thread running through all five takeaways, from Walid and Stewart’s conversation to our customers’ hard-won lessons to what my team built: your AI is only as good as your data.  

A solid data foundation and a fast path to building on it belong in the same plan. If you missed the event or want to revisit a session, watch Precisely Now on demand to see the Precisely Platform, Precisely AI Studio, and the honest conversations for yourself.