“Context is King,” Says Chris Peart, Snowflake

June 9, 2026

Chris Peart, Sales Leader at Snowflake Canada, shares how unified data, governed context, and agentic AI are reshaping how enterprises turn information into action. He explains why “context is king” as frontier models become commoditized, how a single AI Data Cloud across AWS, Azure, and GCP removes the brittleness of traditional architectures, and how Snowflake Cortex and CoWork give knowledge workers immediate answers instead of waiting weeks for engineering teams.

He also digs into the agentic future and the cultural shift required to win with AI: why “nobody sells anybody anything” and customers buy outcomes, why Canadian enterprises are falling behind global peers by being too cautious, and how the next frontier is autonomous agents negotiating with other agents across organiational boundaries.

The episode answers: 

• Why is your data, not the model you choose, the true competitive differentiator in the agentic era 
• What does a governed context layer look like, and why is it the foundation for agents you can trust 
• Why are Canadian enterprises falling behind global peers on AI, and what does it take to start swinging? 

Watch the interview:

Read the podcast as an interview:

(The interview was shortened and edited using ChatGPT)

Mark Kohout: Hello and welcome to this Adastra podcast. My name is Mark Kohout, and I lead the Data and AI Governance practice at Adastra North America. Adastra is a global data, AI, and cloud systems integrator. Today’s episode focuses on the cloud-agnostic Snowflake data platform. Joining me in our Toronto studio to talk about how Snowflake helps customers transform their business with data and AI is Chris Peart. Chris is a Sales Leader at Snowflake Canada, where he oversees strategic enterprise customers across Eastern Canada. Before Snowflake, he held roles at SAS, Salesforce, and Intelex. Chris is based in Toronto and is a Wilfrid Laurier grad. Chris, thanks for joining us today. 

Chris Peart: Nice to be here, Mark. I’m looking forward to it. 

Mark Kohout: Let’s start with you. You’re a founding member of Snowflake Canada back in 2018. Take us back to those early days. What was the pitch when you were introducing Snowflake to Canadian enterprises for the first time? 

Chris Peart: It’s incredible to think Snowflake has been active in Canada since 2018. The message at that time wasn’t too dissimilar to what we’re seeing now. We weren’t talking about AI in 2018, but we were talking about uniting data, getting rid of data silos, and allowing people to get more intelligence and action out of their data by mobilizing it in one central data platform. 

Mark Kohout: A single source of truth. 

Chris Peart: That’s right, across AWS, Azure, and GCP. 

Mark Kohout: And that holds true today. 

Chris Peart: It does. That message has evolved to a point where it’s still that single source of truth, but now that truth holds the context your AI agents and AI projects need to be effective in production, not just in pilot phase. 

Mark Kohout: So you’re folding AI into the platform. 

Chris Peart: AI has always been part of the Snowflake platform to a degree, and now it’s just taking a bigger role. 

Mark Kohout: If I’m not mistaken, you’ve helped Snowflake scale to something like $100 million in annual product revenues. 

Chris Peart: I’m not an accountant, so I can’t verify those numbers, but we’ve done great business in Canada. We’ve helped companies see value out of their data, from small mom-and-pop shops to the largest banks, insurance companies, retailers, manufacturers, and aviation companies. They all have different data and AI needs, but they can get value from a single, governed, unified data platform in Snowflake. 

Mark Kohout: It speaks to the flexibility and scalability of the platform. What were some of the most pivotal moments or challenges along that journey from 2018 to today? 

Chris Peart: One of the most challenging things was helping people see Snowflake for what we are. We’re a new way of looking at data. No longer do you have to amass data in a data lake and then use ETL to move data between data lakes and data warehouses. That introduces complexity, increases cost, and becomes brittle. We needed to prove that what we were doing inside Snowflake wasn’t magic. It felt like magic, but it was about taking advantage of what AWS, Azure, and GCP allow Snowflake to do as phenomenal partners. That’s getting insight from data faster, and for customers to wrap their heads around that took a minute. But the market really sees that now. 

Let me give you an example. There are two types of people. One logs into their data platform at the beginning of the day and has to figure out how to take advantage of a new feature. They want to use a copilot or an LLM, so they need to check their environment, make sure they’re on the most current revision, go through an upgrade cycle, and see what breaks. That’s how people typically interact with software platforms. 

The other side is what Snowflake does. We have one product, just Snowflake, the AI Data Cloud. When customers log in, they can say, “I want to take advantage of Snowflake Cortex,” for example, and once it’s available in private preview, public preview, or GA, it’s just there. There’s no managing upgrades, no worrying about hardware or cutovers. They use it inside their secure, governed platform. 

Mark Kohout: So there’s a control layer to make sure the data is fit for purpose and properly governed. 

Chris Peart: 100%. 

Mark Kohout: Let’s turn to enabling business transformation with AI. Can you explain the relationship between Snowflake CoWork and Cortex? 

Chris Peart: Snowflake CoWork is the knowledge worker’s interaction with Snowflake. 

Mark Kohout: An interface. 

Chris Peart: It’s a clean front end. It allows people to ask questions and get answers back very quickly. Cortex is the brain that sits underneath. Cortex does the appropriate routing. When a knowledge worker asks a question, Cortex structures a SQL command that retrieves the data, or if the data lives in unstructured format, Cortex uses vector embeddings to pull data from that unstructured source. The goal is to increase knowledge workers’ speed to insight by giving them answers in one simple chat interface. 

Mark Kohout: That’s a real manifestation of fast time-to-market and speed-to-insight. When a business user sits down with CoWork for the first time, what can they do that they couldn’t do before? 

Chris Peart: Every knowledge worker listening probably has beautiful dashboards. The idea is to have dashboards and reports that answer all the potential questions a business user may have. But that’s an impossible goal. As a knowledge worker, every time you get an answer, whether a chart, dashboard, or report, you’ll probably have three to five follow-up questions. Traditionally, that requires emailing a data team to pull the data, develop a pipeline, cleanse it, and email it back. By the time you get the answer, you’ve forgotten why you asked. 

Mark Kohout: Sounds like a data delivery bottleneck waiting to happen. 

Chris Peart: 100%. With Snowflake CoWork, you can ask those follow-up questions and get answers immediately, so you can test hypotheses and react quickly. 

Mark Kohout: So you can follow business intuition and curiosity as it comes to you. 

Chris Peart: Think about a company called Fanatics, an unbelievable global sports brand. They do sports marketing and merchandising for F1, NBA, tennis… 

Mark Kohout: The majors. 

Chris Peart: And the NHL. They collect 2 billion signals every day about what individual sports consumers like and don’t like, and how to take advantage of that attention. We’re here in Toronto, so Fanatics knows how best to market to me. We’re recording during the Eastern and Western Conference finals in hockey. They know it’s my nightmare to potentially see the Habs in the Stanley Cup finals playing against Mitch Marner’s Vegas Golden Knights. They understand how conditions change for individual consumers and how to maximize revenue from me, or from the entire GTA, as we sit there in horror. 

Mark Kohout: By the way, go Habs. 

Chris Peart: I’m okay with the Habs. 

Mark Kohout: There’s a thread running through what you’re telling us, Chris. It’s about agility, leveraging data and AI as a strategic asset, and finding competitive advantage through the ability to react. Any other use cases that made you think, “This is why we built this”? 

Chris Peart: Absolutely. There’s a Canadian customer, an unbelievable cybersecurity company called eSentire. They collect 25 petabytes of cybersecurity, network traffic, and infrastructure data daily. They give that to their threat detection agents to understand whether nefarious actors are inside or trying to breach a network. They’re doing this at scale for significantly less money than before, which means they can handle even more detections for the betterment of their customers’ security. 

Mark Kohout: That’s an incredible footprint. Have you seen customers reporting tangible, measurable results? That’s always top of mind for business leaders, but it can get fluffy. 

Chris Peart: To get measurable results, you need to be measuring the current state. A lot of people want measurable results from AI without fully understanding their current state. My team works to understand the current state so we can point to defensible metrics, like increased revenue, reduced costs, and reduced risk, together with the customer. 

Here’s an example. Siemens Energy, a massive energy manufacturing company that makes turbines and large machines, had over 800,000 technical documents that new knowledge workers needed to learn to be effective. It actually took a new knowledge worker about seven years of non-stop reading to get that information into their minds, let alone retain it. They loaded all that unstructured document text into Snowflake and use CoWork so new knowledge workers fixing machines or developing new turbines can leverage that institutional knowledge to solve problems faster. 

Mark Kohout: There’s an old MBA saying about knowledge management: “Siemens doesn’t know what Siemens knows.” Now, with Snowflake, Siemens knows what Siemens knows. 

Chris Peart: I’ve never heard that before, Mark, but I love it. 

Mark Kohout: I’ll send you the bill for the royalties. What you described with Siemens is also germane to AI-driven transformation. As data managers, we often focus on structured and semi-structured data, but so much knowledge sits in documents that can be completely dark. Now everyone’s talking about agentic AI. We were getting our heads around LLMs, but the autonomy provided by agents is shifting the game. How is Snowflake thinking about agents and autonomous workflows? 

Chris Peart: Gen AI is your ability to get answers to questions and ask why. Agentic AI gives your agent the ability to perform work based on those answers without human intervention. The way we think about it is: context is king. In a world where frontier models are continually leapfrogging one another, including Anthropic, OpenAI, Gemini, Grok, plus open-source and Chinese models, you’re going to see commoditization of the models themselves. 

Mark Kohout: Rather than a winner-take-all game. 

Chris Peart: In a world where anybody can use the best model possible, your competitive differentiator is going to be your data, the context you have. With agents potentially working with other agents autonomously, your ability to have secure, governed context in that semantic layer, accessible through AWS, Azure, and GCP, is going to allow you to move faster and get more value than your competitors. 

Mark Kohout: What platform functionalities enable these agentic operations? 

Chris Peart: From a Snowflake perspective, you can manage orchestration through Snowflake CoWork and Snowflake Cortex Code. We’ve shortened that to Snowflake CoCo. You can launch commands and use that same question-answer-action interaction to work with MCP servers in other applications, like Google, Workday, Salesforce, ServiceNow, and SAP, to perform actions. 

I think there’s a world coming where a supply chain agent at a retailer notices a potential product shortfall and autonomously negotiates a purchase agreement with a sales agent at a vendor. I’m not aware of anybody fully there yet, but I see it on the horizon. In that reality, if your data isn’t in one place, governed and secure, with that context layer… 

Mark Kohout: It’ll break. 

Chris Peart: It’ll break. The people who take advantage of this are the ones who will win. 

Mark Kohout: Fascinating. Moving from human-in-the-loop to human-above-the-loop, supervising. Let’s turn to relationships with customers. Your career spans SAS, Salesforce, and your success at Snowflake. What have you learned about making a great customer relationship? 

Chris Peart: Honestly, it comes down to trust. Nobody is going to sell anybody anything. Customers will buy software; they’ll buy outcomes. I’m not in a position to magically influence somebody to do something they’re not willing to do. The customer on the other side doesn’t buy software as often as I sell it. They’re putting their career and reputation at risk, making a bet. They need to understand that the person sitting across from them, my team and I, have their personal and professional best interests at heart. There are a lot of ways to do that, but it comes down to ensuring that if they work with us, it’s going to be beneficial for them both personally and professionally at the end of the transaction. 

Mark Kohout: It sounds like you become an extended member of their team, almost a trusted advisor. 

Chris Peart: We try, and we try to earn that trust every single day. When they spend a dollar on Snowflake or with my team and me, we want them to get $10 in return, because then they’ll willingly want to spend $10 to get $100 in return. 

Mark Kohout: What can we do next? 

Chris Peart: Exactly. Let’s keep that going so they’re constantly getting outsized value for the relationship and the investments they’re putting into the platform. 

Mark Kohout: Now that AI has become so central to the Snowflake story, how are these customer conversations changing? 

Chris Peart: We’re able to showcase how Snowflake itself uses AI. 

Mark Kohout: Walking the talk. 

Chris Peart: 100%. We show up to meetings incredibly well prepared because we use Snowflake CoWork to understand customers’ pain points and the problems they’re trying to solve, synthesizing a vast amount of knowledge into a conversation. That’s where we start to show them how we can help drive that change inside their own organizations in 90 to 180 day sprints. 

Mark Kohout: Does this apply to organizations of all sizes? You mentioned mom-and-pop shops and S&P 500 organizations. Do you see patterns in who’s getting real value, who’s extracting the most value from data and AI platform investments? 

Chris Peart: That’s a good question. For me, it’s the folks willing to take a risk and willing to fail and learn from that failure. Life is full of failure. It’s impossible to get through life and business without it. I have lots of meetings and clients that don’t work out. That happens. But the folks who endure that failure and learn from it are the ones getting value fastest. 

In Canada, I’m Canadian, born and raised in Toronto, and we tend to be more cautious than the rest of the world. We point to regulatory requirements and other issues to justify our caution. I want to challenge all the Canadians listening to this podcast: our peers with the same regulatory requirements and the same issues, whether in the UK, Australia, New Zealand, or down south in the US, are taking big swings and winning with AI faster than we are. I’d like to see us turn that narrative around. The folks willing to take that risk are the ones winning. It’s scary, I get it, but there are ways we can accelerate positive outcomes by working with Snowflake and the team at Snowflake. 

Mark Kohout: Perhaps a key to that increased risk appetite, the willingness to put yourself out there and fail fast, is also understanding the risks and how to mitigate them. If you’re going to fail, fail smart, and do that learning. Many of the controls we’ve talked about within a platform like Snowflake are key to that. 

Chris Peart: And that’s the big thing with Snowflake. To touch on what we’re doing differently, it’s one single platform available through AWS, Azure, and GCP. You don’t have to manage multiple instances or accounts. It’s one interface that’s secure, governed, and trusted. Even when you’re using gen AI and Snowflake CoWork, you have access to data based on your security posture. 

Mark Kohout: Through RBAC. 

Chris Peart: Right, role-based access control. You’re never seeing anything you’re not supposed to see, like merger and acquisition documents or salaries. You, as the person interacting with the AI or the agents working with you, all inherit the same security and visibility permissions you have, so they’re never doing anything nefarious or outside the realm of your role within the organization. 

Mark Kohout: Chris, you just stole my thunder. My last question was exactly that. It sounds like you’ve made a compelling case that Snowflake is the right place to start. For data and business leaders in our audience who are still on the fence about getting started, what’s your advice? 

Chris Peart: It’s easy. Go to ai.snowflake.com. There’s a free trial. You can get started with just yourself and your dog and a credit card, or you can scale up to a Fortune 500 company. It’s the same experience and platform either way. The goal is to ensure people are getting value out of their data. 

Mark Kohout: Just make sure the dog doesn’t eat the credit card and you’re good to go. Chris, we’re out of time. Thank you so much for taking time out of your busy week to share your insights on cloud-enabled AI and how Snowflake helps your clients make it happen. I’m taking away messages about agility, that you can get going faster than you thought possible, and that you can do it in a controlled way with unified data, by pushing AI tools right into the platform. Thanks again for being with us. 

Chris Peart: Thanks for being great partners. Really appreciate it, Mark. 

Mark Kohout: And to our audience, if you’ve enjoyed today’s discussion, be sure to like and subscribe to this podcast series for more insights on data and AI led business transformation. Until next time, thanks for listening, and so long for now.