Here’s a nerdy post about why I joined Snowflake after nine years at Amazon. I hope it’ll help those in similar situations. Let’s rewind time to July when I had to make the call - to stay at Amazon or join Snowflake. Let me use Jeff Bezos’ Regret Minimization Framework.

Opportunity and Regret

If I stay at Amazon, I’ll regret missing opportunities at Snowflake. If I move and join Snowflake, I’ll regret leaving Amazon opportunities behind. The bigger the opportunity, the bigger the regret. I promised nerdy: I see opportunity as the mathematical product of the value one can gain and the probability of such outcome. One can express the value in salary, prestige, satisfaction, excitement, or whatever matters to them. Here’s what mattered to me.

If I Join Snowflake…

I’ll have a chance to transform cloud computing for data-intensive applications. I’ll work on Native Applications which may become the de facto platform and marketplace for data apps, think AppStore. I’ll join a relatively small and fast-growing company with much more growth ahead as it expands into the app territory. I’ll work with great people who joined from Google, Apple, or Meta. I’ll hire many top-notch engineers given the top-notch compensation I’ll have to offer.

If I Stay at Amazon…

I may get promoted to senior manager, I think I’m on track. I’ll work on exciting initiatives to bring Aurora Serverless to the next level. I’ll feel comfortable knowing the product. I’ll have support from the network I’ve built at Amazon over the years. I’ll keep being challenged by AWS’s scale and its leaders, industry’s best.

A Hard Choice - Stay at Amazon or Join Snowflake?

The large upswing of the Snowflake opportunity tips the scales. The value of Snowflake is so high it compensates for uncertainty. The mathematical expected value is higher. Even disregarding the hard value, I won’t regret the Amazon stuff that much, especially the title. I’ll surely regret not being part of Aurora anymore but still, I’ll cheer the team from the sidelines and will continue seeing Aurora Serverless as my baby. Snowflake it is!

Outcome

It has been three months since joining Snowflake and the opportunities I bet on in July are starting to feel more real. Native Applications are a bigger deal for Snowflake than I anticipated, with lots of organizational support. Even Benoit and Thierry, our co-founders, are hands-on with Apps despite a ton of other responsibilities they have. That’s a big, positive surprise to me. After these three months, I already feel welcomed and valued at Snowflake. I couldn’t have imagined a better start and brighter future.