Keith Belanger on How to Build Trusted Data Products for Enterprise Teams

What do you choose — speed or control — when building enterprise data products? Is it possible to have both?

In this conversation, Keith Belanger, Field CTO at DataOps.live, unpacks what it takes to engineer data solutions that deliver speed, scale, and trust, without compromising one for the other.

With a career spanning architecture, governance, and AI/ML, how does your experience inform your decision-making as Field CTO?

My career has always lived at the intersection of business and data strategies. I’ve led enterprise data initiatives both in the operational and analytical space, implemented data governance programs grounded in people and process, and supported modernization efforts across multiple waves of change, each redefining what “modern” means in the data space. But through it all, one thing has remained clear. The real differentiator is how well your data solutions align with your business goals.

To help organizations meet this objective, I’ve always taken a strong design-first approach, rooted in data modeling and best practices, rather than rushing to build in an ad hoc way. Standards and governance matter, but so does autonomy. Great practices give teams the freedom to move fast while still working within a trusted, consistent framework.

As Field CTO, I apply these lessons daily to help ensure our platform delivers more than just technical proficiency. We enable the rapid delivery of trusted data products. Every feature we prioritize is built to support operational resilience, governance, and testing. Because in the end, you need to operate at the speed of your business, or your value will be lost.

Many platforms claim to offer "end-to-end" DataOps. What should data teams look for when evaluating solutions that go beyond marketing promises?

“End-to-end” should mean more than a buzzword. If a solution claims to deliver end-to-end impact, it must touch every part of the data lifecycle, horizontally across ingestion, transformation, delivery to data consumers, and vertically through the development lifecycle — from development environments all the way to production deployment.

But it doesn’t stop there. True end-to-end DataOps requires orchestration, auditing, governance, testing, and seamless integration. And not just within a single tool, but across the broader ecosystem of third-party solutions like ETL/ELT platforms, data governance solutions, data quality solutions, and BI applications. End-to-end orchestration means coordinating all of these components cohesively. That’s why we advocate for the Seven Pillars of True DataOps, a foundational framework that includes:

1. ELT (and the spirit of ELT)
2. Continuous Integration / Continuous Delivery (CI/CD)
3. Component Design and Maintainability
4. Environment Management
5. Governance and Change Control
6. Automated Data Testing and Monitoring
7. Collaboration and Self-Service

These aren’t optional, they’re essential for delivering reliable, scalable data products. In my experience, one of the biggest challenges has always been getting different solutions to work together cohesively. It’s not just about Snowflake as the data platform; it’s about how your ingestion, transformation, governance, and testing layers all interoperate. That’s what creates a harmonious, well-orchestrated system. When evaluating platforms, teams should look for solutions that support these seven pillars and ultimately choose those that help deliver trusted data at scale, not just automate for automation’s sake.

What new capabilities does the latest DataOps.live – Snowflake partnership bring to data engineering teams?

This latest partnership was born from Snowflake identifying key friction points that their customers were experiencing, specifically around CI/CD and operationalizing their dbt projects in an enterprise setting. They recognized that DataOps.live had solutions to address these challenges and wanted to provide a frictionless experience to their customers by embedding those capabilities directly into their Snowflake Data Cloud accounts.

They also wanted to eliminate long procurement cycles, and ensure there were no independent infrastructure to manage, and complex integrations to implement. Rather, they wanted native capabilities that customers can access and use as easily as they would Snowflake features like Snowpipe, Dynamic Tables, or Cortex.

What resulted was the introduction of our Dynamic Suite, which includes two Snowflake native apps from DataOps.live:

• Dynamic Delivery – CI/CD for Snowflake
• Dynamic Transformation – Operationalize your dbt™ projects

Both apps are available today in the Snowflake Marketplace at no cost for the first 500 minutes per month, and then a pay-as-you-grow model, just like Snowflake customers are already used to.

What makes this even more powerful is that these apps are not just isolated point solutions. They’re on-ramps into the broader enterprise DataOps.live Platform, allowing organizations to scale their data operations over time, all within a single, consistent user experience. Better yet, this is just the beginning as we're actively working with Snowflake’s product team to introduce additional Dynamic Apps aligned to their product roadmap.

Sitting at the intersection of Snowflake teams, enterprise customers, and internal stakeholders, how do you lead product strategy in a way that aligns with all three?

It starts with active listening. We spend time with customers, solution integrators, Snowflake SEs, and folks in the trenches with Snowflake customers to understand pain points, real-world use cases, and emerging patterns. My role is to help interpret that information, layered with my years of industry experience, and translate it into input and broader go-to-market awareness from a marketing perspective.

We align closely with Snowflake’s product priorities, like CI/CD and dbt™, but we also stay tuned in to the broader capabilities customers want to adopt, like Dynamic Apps, Cortex AI, and Data Engineering in general. Our goal is to ensure that Snowflake innovations can be leveraged in a way that meets enterprise DataOps needs with governance, automation, and testing already built in.

Internally, I help ensure that what we build not only solves technical challenges but also fits the challenges data practitioners face every day. I also reinforce our long-term vision as The Data Products Company making it easier for organizations to deliver trusted data solutions with the speed and control modern enterprises demand.

You’re highly active in the data community as a speaker and content creator. How does that external engagement influence your product and GTM thinking?

Let’s face it, I’m a data architect at heart. Engaging with the data community keeps me grounded in the real world. I still remember being hesitant to share my perspectives publicly years ago, but I’ve found the community to be incredibly welcoming. Not everyone agrees with my viewpoints, and that’s the beauty of it. I’ve learned just as much from the diversity of thought as I have from sharing my own.

I never originally planned to work for a product company, but this is now my second one. A few years ago, I started to see it as an opportunity, thinking maybe my years in the trenches could help shape products in a way that truly serves data practitioners. After all, I did my fair share of venting about products that often missed the boat over the years. Here at DataOps.live, I bring the voice of the customer into the organization and provide perspectives that might otherwise be missed. Whether I’m speaking at a conference, writing a blog, or debating with peers online, I’m constantly getting a pulse on what people are struggling with and what’s inspiring them. That feedback loop directly shapes both my strategy thoughts and our go-to-market approach. In many ways, community conversations are where the future shows up first. I just try to keep listening.

How is DataOps.live positioned to support evolving enterprise needs like data mesh and real-time observability within DataOps workflows?

We don’t ascribe to a single methodology. Instead, we provide a platform that brings True DataOps capabilities, governance, automation, testing, and observability to any approach, regardless of how centralized or decentralized, simple or complex, small or large the environment may be.

That’s the power of our solution. As I mentioned earlier, the Seven Pillars of True DataOps are designed to support flexibility without sacrificing trust or control. And after all the years I have been in the field, I’ve learned that a new “modern” methodology is always around the corner. The strength of a truly great DataOps platform is that it can adapt and layer on top of any data strategy and still deliver consistent, enterprise-grade results.