
Control vs. Automation: Why Light's Bet Is That AI Makes Finance More Accurate, Not Just Faster
The AI Revolution ShowShow Notes
Light co-founder Jonathan Sanders built two companies before this one, migrated both to Oracle ERPs, and decided there had to be a better way. Jonathan shares why legacy ERPs are failing in the AI era, how Light uses agents to check the output of other agents (and why that is more accurate than humans doing the same job), and what it actually takes to get finance professionals to trust AI with the general ledger. In this episode, we explore: (02:10) Why Jonathan founded Light after two painful ERP migrations (06:09) Why legacy ERPs cannot make AI work, and what Microsoft Copilot gets wrong (09:07) How Light uses AI to lower switching costs and win rip-and-replace deals (10:46) Speed, control, and accuracy: the three things finance teams actually need (14:09) How Light outperformed humans plus machines on bill parsing (21:29) The memory capability inside Light that learns your business from transactions alone (25:41) How a 100-person company runs its entire finance function with one person (29:08) Finance hackathons with JP Morgan and Deloitte, and what surprised every attendee (32:00) Why an immutable general ledger is the foundation of trust in agentic finance
The AI Revolution Show is the intelligence podcast for B2B software founders navigating the platform shift, the founders building through it, and the founders already on the other side. Hosted by Alex Theuma, founder of Shift AI Events. Subscribe so you don't miss the next episode.
More from The AI Revolution Show
Only 1% of website visitors ever fill in your form. Arjun Pillai built a company around the other 99%. In this episode, Alex Theuma sits down with Arjun, founder and CEO of Docket, an AI demand agent serving enterprise and mid-market B2B companies, to talk about what happens when buyers stop trusting the tools that were supposed to help them grow. Arjun shares how Docket signed its first eight design partners from nineteen cold conversations, why outcome-based pricing does not work for most GTM use cases, and what a "marketing factory" actually looks like in practice.
What do you do when your bootstrapped business hits $5M ARR but stops acquiring new customers entirely? Alex Turnbull didn't wait for the answer. He shut it down and started over. Twice. Host Alex Theuma speaks with Alex Turnbull, CEO of Helply, formerly CEO of Groove for 15 years. With 100 companies already committed with deposits and a $165K MRR pipeline before launch, Turnbull explains why he killed two products to build one agentic-first platform, no seats, no legacy baggage.
Finance teams at Anthropic, Coca-Cola, and Unilever have stopped using spreadsheets entirely. Is your business planning keeping pace with the speed of the market, or are you running a Formula 1 race with a bicycle? Alex Theuma sits down with Eléonore Crespo, co-CEO and co-founder of Pigment, the AI performance management platform that has doubled revenue for three consecutive years and recently crossed $100M ARR. Eleanor shares why bolt-on AI is a losing strategy, how Pigment is cutting enterprise implementation from nine months to 20 days, and what it actually takes to build and defend a moat in the age of agents. In this episode, we explore: (02:06) Why Eleanor keeps engineering in Europe and sales in the US (04:09) Crossing $100M ARR with three straight years of revenue doubling (08:11) Why large enterprises cannot just run their business on Claude (10:23) How Pigment pivoted from pre-AI infrastructure to agentic workflows (12:05) Eleanor's three rules: AI-native thinking, moat-building, and speed (22:50) Why two co-CEOs works, and how they handle opposing visions (27:31) Launching a free trial inside an enterprise sales motion (33:10) What people dangerously underestimate about execution and great teams
AI didn't break SaaS. It just made the cracks impossible to ignore. And according to Jacco van der Kooij, the AI industry is now sprinting toward the exact same cliff edge, only faster. Alex Theuma sits down with Jacco van der Kooij, founder of Winning by Design and advisor to companies including Canva, SpaceX, and Boston Dynamics. They cover why inbound-dependent growth models are a ticking clock, where AI is being wasted on old processes, and why human-to-human connection is still the dominant force in sales. In this episode, we explore: (01:32) Jacco on what excites and scares him about AI right now (03:29) Why AI exposed SaaS problems rather than causing them (05:09) Using AI to eliminate your top six business constraints (06:13) The Ferrari-as-lawnmower mistake most founders are making (17:57) High-touch vs. high-volume sales: where humans still win (21:27) Why applying AI to old processes changes nothing (22:11) The inbound dependency trap, and why AI companies are repeating it (23:41) What Jacco expects to learn, not just teach, at Shift AI Europe The AI Revolution Show is the intelligence podcast for B2B software founders navigating the platform shift, the founders building through it, and the founders already on the other side. Hosted by Alex Theuma, founder of Shift AI Events. Subscribe so you don't miss the next episode.
Alex Theuma sits down with St John Dalgleish, CEO of Perlon AI, to talk about what actually separates a working AI GTM motion from a wasted one. They cover why ownership matters more than tooling, where the human salesperson still wins, and why AI cold callers are a brand risk few founders are willing to take. In this episode, we explore: (01:36) Why most companies are still failing with AI SDRs (05:29) The ownership problem inside sales teams using AI (08:02) How Perlon AI qualifies customers who are actually ready (09:57) From consulting spinout to seven-figure ARR product (12:16) Where the human SDR still beats automation every time (14:34) St John on why AI cold callers are a brand liability (19:23) Dogfooding at scale: 70% of pipeline from their own tool (22:07) The AI backlash driving a comeback for face-to-face events (31:35) Three things Perlan AI actually differentiates on in a crowded market