Clektive Thinking: Snowflake Summit 2026
For the first time, the energy at Snowflake Summit 2026 wasn’t around what AI could theoretically deliver, it was about what AI is genuinely delivering right now, in production, for real organisations.
In this episode of Clektive Thinking, Chris and Elliott break down what the announcements mean for business leaders ready to act.
CoWork, CoCo, and the new Snowflake paradigm
Two of the most talked-about announcements centred on how Snowflake is repositioning its tooling around two distinct types of users. The first is the builder — the data engineer, developer, or technical analyst who works inside the platform, constructing data models, agents, and applications. The second, and newer, concept is the knowledge worker: the executive, sales leader, HR professional, or anyone else who needs to use and act on data without wanting to build anything.
To serve this knowledge worker, Snowflake has introduced CoWork — a repackaging and evolution of what was previously called Snowflake Intelligence. CoWork lets business users interact with their data directly from desktop or mobile, generate dashboards, create artefacts, and schedule recurring queries. Alongside that, Cortex Code has returned to its original project name, CoCo, signalling a renewed focus on code-centric workflows for technical users.
“If I’m a CFO and I need to put commentary on last week’s trading, I can use CoWork to schedule that. Because it’s been built with my data and my semantics, I’m spending time reading and taking action — not wrangling a spreadsheet.”
— Elliott Fairhall, Clekt
What does ‘agentic enterprise’ actually mean?
Agentic AI gets used a lot right now, and it can start to feel like noise. But there is a real and meaningful distinction worth understanding. Generative AI answers questions. Agentic AI takes action – and keeps going.
In Snowflake’s vision of the agentic enterprise, organisations aren’t just querying their data to get an answer. They’re setting up agents that continuously monitor information, spot what’s going well and what isn’t, and take meaningful action in response without requiring a human to be in the loop at every step. You configure the work, and it runs. When you come back, the task is done.
The practical applications here are genuinely compelling. Automated board packs, real-time trading commentary, meeting summaries, communication drafts. These are all tasks that once consumed hours of manual effort becoming something the system handles. And because everything is grounded in an organisation’s own data and agreed semantics, the outputs are trusted and actionable, not generic.
The Clekt perspective: At Clekt, we call this Assisted Intelligence — and it’s the lens through which we think about AI at work. The goal isn’t to replace the people in your business. It’s to free them from the tasks that shouldn’t need human time, so they can focus on the thinking, judgement, and creativity that genuinely does. The agentic enterprise is a version of that at scale.
Are businesses actually ready for this?
The honest answer is: some are, and many aren’t — yet. And that’s okay, as long as organisations are clear-eyed about where they stand and what steps come next.
The most significant insight from Summit wasn’t about the newest feature announcement. It was a simple statistic that grounded everything: 85% of enterprise clients arrive at Snowflake with a data problem before they have an AI problem. That’s the real starting point for most organisations. Before you can build a trustworthy agentic enterprise, you need data that’s well-governed, semantically rich, and genuinely understood across your teams.
“You don’t want to build your agentic enterprise on fragmented, ungoverned, half-trusted data,” Elliott put it plainly. “You want to feel confident. You want to know that if an agent is looking at profit, it’s using the same definition of profit that your CFO uses.”
The question has shifted, and so should yours
The most useful reframe to take away from Summit is this: the question is no longer what can I do with AI? The possibilities are, for most practical purposes, vast. The question that matters now is how do I get my data, my governance, and my organisation set up to actually run with this?
That means starting with outcomes and being clear about what you’re trying to achieve before reaching for the technology. It means building a semantic layer that reflects how your business actually thinks about its data. It means working collaboratively across departments and the C-suite to agree on priorities.
And it means treating governance not as a brake on progress, but as the foundation that makes everything else possible.
Three things worth doing now
You don’t need to solve everything at once. The organisations making the most meaningful progress aren’t the ones who did it all in one go — they’re the ones who started somewhere deliberate.
- Get clear on your business use cases first. CoWork and agentic tooling exist right now. But if you’re not confident in your data or your semantic layer, use these tools initially to pressure-test your working assumptions rather than going straight to production. Let the technology help you validate before you scale.
- Define what good data looks like for your organisation. Every business will have access to the same frontier models e.g. OpenAI, Anthropic, Google, all available within Snowflake’s platform. The differentiator will be your data: how well it’s organised, governed, and understood. That’s the competitive edge worth building now.
- Work with people who can help you move with confidence. The opportunity is real, but so are the risks of getting it wrong. Whether that means validating your data foundation, building your semantic layer, or designing your first use cases, the best approach is collaborative and incremental and building a plan that works across your business, not just within a single team.
Clektive Thinking is Clekt’s podcast series exploring data, AI, and what it means to put technology to work for real teams. To find out more about how Clekt can support your data and AI journey, get in touch using the contact form below.
