Finding where AI actually helped a deal team
A private equity deal team wanted to use AI but didn't know where it would help. I spent two weeks working their hours, found the bottleneck in the time data, and built the roadmap around it. Review time dropped 17%.
- Organization
- EY, for a private equity deal team
- My role
- Senior Technology Consultant, AI Strategy
- When
- 2024 to 2025
- Result
- Deal review time down 17%; proposal win rate up 20% the following year
The problem
The deal team had a general interest in using AI but no idea where to start, and real doubt about whether automation would add anything or just be a gimmick. My job was to turn that into a roadmap aimed at an actual bottleneck.
What I did
I didn’t know due diligence workflows, so I flew to Atlanta and shadowed the team for two weeks of their hours, which ran around 90 a week. I mapped how their days actually went and pulled their time tracking data to see where senior people were getting stuck.
The data pointed at the first stretch of a deal: gathering and reviewing documents. That’s where the most expensive time was going, so I built a phased roadmap around it and prioritized a pilot to speed up document review. I ran workshops with leadership on the tradeoffs and worked between the developers and the deal partners to turn what the partners needed into technical specs. Before anything rolled out, we tested the tool against historical deals to get real accuracy numbers.
Result
The pilot data justified the full rollout. Deal review time dropped 17%, and the team won 20% more proposals the following year.
What I took from it
When the goal is vague, the fastest way to a good answer is to watch how people actually work and look at the data before deciding what to build.