**The big picture:** Deel's SVP of Product, Aaron Goldsmid, details how the company found over $140 million in annual recurring revenue (ARR) by automating previously overlooked, small manual tasks, equivalent to 600 people's work. **Why it matters:** This case study highlights a significant shift in how companies can leverage AI for efficiency, challenging traditional headcount as a measure of productivity and emphasizing 'leverage' instead. **Between the lines:** - Deel achieved $140M ARR by automating numerous small, unglamorous tasks that were previously deemed not worth an engineer's time. - Aaron Goldsmid, SVP of Product, states that headcount is no longer a useful proxy for output, advocating for measuring individual leverage. - Deel built a 'moat' by training small language models on its proprietary data corpus, achieving frontier-model quality at a fraction of the cost. **Staffing & HR impact:** This approach suggests a future where HR and staffing leaders must re-evaluate talent metrics, focusing on the impact and leverage of a smaller, highly efficient workforce rather than just headcount. It also underscores the potential for AI to significantly boost operational margins through automation. **The bottom line:** The true ROI of AI lies in automating the 'change on the floor' – the vast backlog of tiny inefficiencies that, when compounded, yield massive returns and reshape workforce planning.