Measuring AI ROI & Orchestrating Token Spend
In 2025 Uber burned through its entire 2026 AI budget in roughly four months and responded by capping employees at $1,500 per month per tool. Amazon shut down an internal leaderboard that ranked engineers by token consumption, with the message "don't use AI just to use AI" after staff gamed the metric. Microsoft, after giving thousands of engineers Claude Code, exhausted its annual budget ahead of schedule and steered teams toward cheaper tooling.
Going into 2026 we’re all facing growing token bills with poor visibility into ROI.
This dinner brings together operators wrestling with the same questions: how to measure engineering output when so much of it is AI-assisted, how to actually quantify the return on every dollar of token spend, and where the emerging tooling layer of model routers and token orchestrators fits in.
Join us in San Francisco for an exclusive dinner where we'll ground the conversation in what the top 1% of SF teams are doing in practice, with hard-won lessons from the likes of Robinhood, Ripple, and others.
Key discussion topics include:
- How do you measure engineering output in the AI era?
- How do you quantify the ROI of token spend?
- Where do model routers and token orchestrators fit in?
- What are the top 1% of teams in SF actually doing (Robinhood, Ripple, and more)?
This invite-only event is designed to create an intimate space for you to meet with a small group of engineering leaders to discuss challenges, exchange best practices, expand your professional network and more.
We hope to see you there!
Questions? Contact [email protected]
