Yesterday Oracle shared that they're continuing to invest in Cohere and Meta's Llama 2 🦙🦙 as model options for their customers. Why? Smaller models allow you to have more control (e.g., increased fine-tuning for particular domains), more data security, and they're cheaper to run. Model selection - It's a topic that become more popular due to hashtag#genai , but it's something that should be discussed more frequently.

In the early days of Uber, I was often the only woman in the room of 10+ executives in late night "jam sessions" strategizing on hashtag#pricing and hashtag#procurement of driver supply hours. This was back at 706 Mission. While it was certainly a privilege to have that experience, I wish I had a group of professional women to connect with and help accelerate my professional development.

Are you incorporating AI tools into your workflow? (and what company isn't?!) Make sure: 1) The AI tool serves your use case 2) The model was trained with legitimate data in an appropriate way 3) Your finance team can forecasts the costs

What do CFOs hate most? Lack of predictability. In my conversation with Jake Gord earlier this week, he shared this UnitedHealth story with me, and I couldn’t believe it. In the world of software, bugs 🪲 exist and outages happen. Our teams work as quickly as possible to fix them.

How is this sustainable for anyone? Could you imagine going into the financial model to forecast the impact of this minimum wage rule? =if(find("bread", Restaurant Name), MinWage = $20, $16) 🍞I bet we will see bread being sold at many more restaurants now.

CEO & CFO: We need to track 1,000,000,000 different KPIs every week. Please report by Monday AM. Finance and Procurement: 🙄 Can we get some automation and tooling to help with that? CEO & CFO: 🦗 🦗 🦗 -- How many of you have experienced this? Two things I've done to help thread the needle: