Nicolai Tangen, CEO of the Norwegian Sovereign Wealth Fund, sits down with Sridhar Ramaswamy, CEO of Snowflake, in this Friday short-form episode. Snowflake is a cloud-based analytic data platform used by roughly half of the addressable Global 2000 companies, powering everything from bank loan approvals to hospital patient data systems.

The conversation covers how AI and coding agents are transforming software engineering and enterprise workflows, the practical barriers that messy legacy data creates for AI adoption, and the mixed legacy of GDPR for European businesses. Sridhar also reflects on growing up in Tamil Nadu in a lower-middle-class household where education and hard work were the defining values instilled by his parents, neither of whom attended college.

What Snowflake Actually Does at Enterprise Scale

Snowflake is described as a cloud computing platform like AWS but with a strong focus on data — ingesting, analyzing, and routing data to action systems.

The Norwegian Sovereign Wealth Fund alone stores 2 petabytes of data on Snowflake and runs approximately 3 million queries per day, illustrating the platform's scale.

Customers span financial services, healthcare, and advertising across more than 25 countries, with roughly half of the addressable Global 2000 as clients.

AI Is Reshaping Sales, Engineering, and Demos

On the sales side, AI agents give sellers instant access to information via their phones, and solution engineers can now build a fully custom demo in 30 minutes with data that looks client-specific.

Spec-driven development is the new frontier: engineers write an English language spec, then automation handles writing, testing, and deploying the first version of the code.

"We have the superstars, people that are 50, 100 times more productive than the average software engineer" — Sridhar, on the gap AI is creating between top and average engineers.

"Software engineering is nothing like what it was 2 years ago" — Sridhar, describing the pace of change in how code is written and shipped.

Agents: Everything Is Programmable and Interconnected

Sridhar defines agents as a model plus code with access to tools it knows how to call intelligently — from writing and executing code to querying a portfolio and returning results.

Snowflake Intelligence can give an agent access to all structured and unstructured data, enabling it to form a plan and answer practically any question without manual data wrangling.

Agents also handle downstream actions: "If you want to send an email based on an analysis you did, you don't have to cut and paste. You can just tell your agent, 'Please send this email to Stefan,' and out goes the email" — Sridhar.

Products like Snowflake's Snowwork concept reflect a broader shift: everything is programmable and interconnected, changing how people fundamentally think about work.

Dirty Data Is No Longer the Blocker It Once Was

Legacy data — full of duplicates, gaps, and decades of stitched-together systems — used to be a huge barrier to AI adoption, but Sridhar says it is "getting simpler by the day."

Adding a single column to a complex data pipeline once required a week of programmer labor; tools like SKILS (English language programs) now automate that entire process in about an hour.

Agent-driven migrations can now move data from legacy systems onto Snowflake in days to a few weeks, compared to the multiple quarters or years such projects historically required.

GDPR: Consumer Rights Won, Startups Lost

Sridhar calls GDPR "a very mixed bag" with significant unintended consequences, noting that regulation must be surgical and that policymakers must rigorously stress-test negative side effects.

The clear positive: consumers can now demand that any company delete all data held about them — a right that forced even Google Ads to build full data-tracking and deletion infrastructure.

The clear negative: GDPR raised the cost of doing business for every European company, and the primary beneficiaries were large incumbents who could afford compliance, while new European startups face the full burden from day one.

The ubiquitous consent pop-ups that users reflexively click through are cited as a textbook example of a well-intentioned rule producing a meaningless outcome.

Tamil Nadu Roots: Education, Hard Work, and Malleability

Sridhar grew up in a lower-middle-class neighborhood in Tamil Nadu and Bengaluru, with a family of four sharing one living room and one bedroom — but with an absolute belief in education as the path forward.

Neither of his parents attended college, yet they were willing to support decisions they were uncomfortable with — including sending their son to a college 300 miles away — which Sridhar frames as a form of intellectual malleability.

He distills his upbringing into three enduring values he now passes on to his own children: the value of education, the value of hard work, and the value of being malleable to a changing world.