Episode Details

Fabian talks with Snowflake CEO Sridhar Ramaswamy about what AI changes for managers, how he hires, how his vision for Snowflake grew and why outcomes are replacing hourly services.

This episode is presented by Minoa. Software companies are moving from selling products to owning their customers’ outcomes, and that only works if you can prove them. Minoa grounds every business case, ROI and use case your team shows a customer in your own data, then measures the value you actually delivered after the sale. Snowflake, Cognite, Vanta and Datadog run on Minoa.

Find out more at unicornbakery.de/minoa

What you'll learn:

Leading teams you don't understand

• "All leaders today are leading teams that they don't understand"
• To stay relevant, managers must relearn what "amazing" looks like in their team's craft
• Unchanged: give credit for wins, take the blame first, be the safety net

Hiring and first-time management

• He hires for intensity, mental agility and team play – not raw intelligence
• His favorite interview question: "Tell me something really hard that you have done"
• Grow autonomy: check-ins every four hours, then daily, then two-week projects

From Neeva to Snowflake CEO

• First AI bets: an AI layer close to the data, unstructured data, text-to-SQL
• His board pitch: a data platform better than the one behind Google Ads
• "A great vision is separated from a good vision based on whether you succeed or not"

Forward deployment and execution

• Customers hate hourly billing: "Customers want outcomes"
• Snowflake's Frontier Engineering Team bids fixed prices for business outcomes
• "I commit, and then I engage" – and why Neeva wasn't different enough

EVERYTHING ABOUT UNICORN BAKERY:
https://zez.am/unicornbakery

More about Sridhar:
LinkedIn: https://www.linkedin.com/in/sridhar-ramaswamy/
Website: https://www.snowflake.com/

Chapters:

(00:00:00) Teaser & intro – manager or founder? "AI makes things very bloody"
(00:03:22) Planning for 9,000 employees – and testing every feature
(00:05:39) "All leaders today are leading teams that they don't understand"
(00:07:13) What hasn't changed – credit, blame and the safety net
(00:08:38) Advice for first-time managers – intensity and adaptability
(00:12:13) Calibrating autonomy – the intern rule
(00:13:59) His favorite hiring question – "tell me something really hard"
(00:17:17) From Neeva to CEO – planning to leave after six months
(00:18:12) Snowflake's first three AI bets
(00:20:34) Coding agents and CoCo – a vision in waves
(00:22:12) The board pitch – a data platform better than Google Ads'
(00:24:13) Good vs. great vision – "the proof of the pudding is in the taste"
(00:27:04) Forward deployed engineers – why software avoided services
(00:30:05) Blurring lines – outcomes over hours
(00:31:57) The Frontier Engineering Team – fixed-price outcomes
(00:33:37) "I commit, then I engage" – and why Neeva didn't work
(00:37:32) Small teams of five or six in the age of AI