Frontier AI, judgment calls, and IPO curveballs (with the SymphonyAI CFO)

Join us as Ravi Narula, CFO of SymphonyAI, walks through 25-plus years leading finance at tech and SaaS companies, including the IPOs of BigBand Networks, Gigamon, and Ooma. Topics covered:

• Whether we should feel reassured or alarmed when frontier AI leaders say the industry needs to slow down (1:18)
• Whether AI can police itself, and what stock option backdating and Sarbanes-Oxley teach about earning trust (4:07)
• What vertical AI looks like in practice, and why SymphonyAI treats the frontier model as the engine and its own product as the vehicle (9:24)
• The 3 layers under the product: decades of industry data, domain knowledge graphs, and customer workflows (11:29)
• How a 2-degree temperature swing at a yogurt plant can shut down a line for 12 hours, and where a 10x ROI on AI comes from (13:59)
• Hard-dollar vs. soft-dollar savings, and cutting false positives in financial crime detection by up to 95% (15:46)
• The one place he won't let AI act alone: irreversible wire payments with no human accountable (17:44)
• Saving an unhappy customer who later helped make IBM Certinia's largest customer, and the 16 CFOs and CAOs who came out of his mentorship (23:03)
• The 1 a.m. discovery a month before Gigamon went effective, and the call that delayed the IPO 6 months (33:21)
• How to raise a hard issue with the board so they actually hear you out (38:42)
• Why an IPO diagnostic covers only a small part of being ready to go public (43:04)
• The 3 p.m. Friday call to Connor Group in 2006 and the 5 p.m. meeting that followed (44:37)
• The time machine question: the lesson on delegating he learned as a new manager at Deloitte (48:43)

For CFOs, Controllers, finance and accounting leaders, and anyone preparing a company for an IPO, this conversation covers the judgment calls no IPO checklist catches, where AI belongs in finance (and where it doesn't), and how good CFOs develop the next generation of finance leaders.

Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.