All Posts

Revenue Recognition · SaaS Finance · AI in Accounting Part 2 of 2 — Deep methodology (~20 minutes) · Quick summary in Part 1 GitHub link here, Part 1 — Quick read: link here Available at SSRN: link Purpose of This Document This piece serves two audiences. The first is the accounting professional who read Part 1 and wants to understand whether the methodology is actually sound, the architecture, the decisions made, the bugs found and fixed, the places where the output require

Revenue Recognition · SaaS Finance · AI in Accounting Part 1 of 2, Quick read (~5 minutes) · Full methodology in Part 2 Full code and output files available on GitHub here and Part 2 link here The Setup I built a complete ASC 606 revenue recognition stack for a fictional SaaS company, 50 customers, 61 contracts, 948 monthly transactions, covering every edge case I could engineer simultaneously: three products with different recognition patterns, five modification types, four

By CA Pushkar Agrawal | The Finance Forge Most payroll projections live in a spreadsheet that only one person fully understands. You have the formula hidden in column AQ, a tab called "DO NOT TOUCH," and a monthly ritual of copying last month's file and praying nothing breaks. I have built those spreadsheets. I have maintained those spreadsheets. And I decided there had to be a better way. So I built a live, interactive payroll projection engine using Python, Streamlit, and A

If you run a business with revenue under $10 million, the phrase "Month-End Close" probably triggers one of two reactions: either you panic, or you ignore it completely until tax season. I get it. You don't have a massive finance team or budget. You likely have one controller, a bookkeeper, or maybe you’re a founder doing it yourself on weekends. The idea of a formal "corporate close" feels overkill. But here is the reality: If you aren't closing your books monthly, you aren'