How AI is turning Figma's finance team into builders: CFO Praveer Melwani
Key points
Key takeaways from an a16z podcast interview with Praveer Melwani, chief financial officer of Figma, hosted by Cack Wilhelm and Yvonne Marcheva (November 2026):
- Finance as builders: Melwani argues the best finance teams are becoming builders — designing systems, data foundations and automation, not just reconciling what happened last month.
- From SQL to AI: When he joined Figma, none of the automation existed; he lived in SQL and Stripe data building golden tables by hand. Today the same work is accessible to any individual through AI tools, from 500-line queries to one-hour answers to what once took weeks.
- Hiring for curiosity: He wants intellectually curious people who keep thinking — AI is a tool to empower you, but you must remain the thinker; his VP finance case studies test how candidates use AI while proving they understand the model they built.
- Automation in practice: A credit and collections lead rebuilt his entire workflow from invoice tracking to Netsuite reconciliation, driving down days sales outstanding; the corporate development team writes investment memos against AI "personas" of each leader to cut cycle time; finance now asks "can you vibe code the product?" as a diligence check.
- Build or buy: A mix — some teams sit in Codex, some build agent skills in ChatGPT, and Figma's own Figma Make lets them build apps in natural language.
- Against slop: In an era where "slop" is being called out, anything you ask someone to consume costs them time and energy; 95% of the challenge is getting your data in the right place.
- 12 to 18 months out: If complexity stayed flat, the team would need fewer people a year from now — automation changes headcount assumptions.
