I spent two years automating myself out of a job. Today the job runs without me ↓ Finsider.ai is live! On every Quality of Earnings I ran, from PwC to 6 months ago, the workflow was the same. 1) Export the books to Excel (1-2 hrs) 2) Build the cash proof by handredactedhrs) 3) Test only bigger numbers, ignore everything else (8 hrs) 4) Keep a 20 page workbook everyone gets lost in or breaks 5) Hope everything moved from Excel into PowerPoint correctly A few companies tried to fix this before me, and I'd used them all. One even got data wrong on a live deal. I demoed 30+ FP&A tools, and they were great for running a finance department, but even with APIs and connectors they couldn't integrate fast enough for M&A PwC tried too and even they couldn't crack it (admittedly I was trying to help it work, but having no AI 7 yrs ago made it hard). So I ultimately built what I needed. Every deal at Petracca Group and Forward Firm over the last 5 years has become an iteration loop. How could I automate more? How could I get better insights with less? Here's what I learned: 1) Many parts of M&A, like the cash proof, can be automated. 2) Sampling is no longer necessary, because machines can scan 100% of transactions. 3) When every adjustment ties to one database on the cloud, the workbook stops breaking. Most of Finsider is plain accounting logic. The numbers either tie or they don't. Even if AI went away tomorrow, 𝑚𝑜𝑠𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑝𝑟𝑜𝑑𝑢𝑐𝑡 𝑤𝑜𝑢𝑙𝑑 𝑠𝑡𝑖𝑙𝑙 𝑤𝑜𝑟𝑘. But the AI really does take it to the next level. I've spent eight years in transaction advisory, starting at PwC, and have run 150+ Quality of Earnings engagements. Finsider is the tool I wanted on every one of them, and as of today anyone can use it. If you know anyone in M&A or looking at financial statements regularly, I'd love to hear from them. Check it out here: redacted Many thanks to everyone who helped me on this journey, two kids and two years later.