Everyone’s being asked what their AI spend produced. We sell the platform that answers that question, so we ran it on ourselves first.
Two people from the same team each built a production system on real company data, and they’re walking through both, including the parts that broke.
Miguel Valdez rebuilt CloudZero’s go-to-market engine, including a Slack bot that routes buying signals and coaches calls, and a Salesforce commissions system built with Claude Code in about six weeks. Emily Allen built the CFO dashboard that reconciles finance data with CloudZero cost data into one view of margin, unit economics, and AI ROI.
Neither build started with a roadmap or an engineering ticket. Both are running now.
You’ll leave able to answer:
- What it actually takes to ship an internal tool with AI, without a dedicated engineering team.
- How to make AI spend defend itself as a percentage of revenue and margin, not a line item nobody can explain.
- Which prompting and steering patterns hold up when finance has to trust the output.
- Where AI didn’t help, and what that says about which projects are worth starting.
Every AI dollar. Every outcome. Connected. This is what that looks like when we do it to ourselves.
