#9: Season Two Starts Here: Reflections, What's Working, and What's Not

#16 - AI Fund

In the Season 2 finale, Anthony Thomas and Zac Henderson are joined by Ellen Li, CFO at AI Fund, the venture studio founded by Andrew Ng that builds AI-native startups from the ground up rather than investing in deal flow. Ellen walks through how her day-to-day has shifted from formatting spreadsheets and manual research to reviewing AI output as she would a capable team member. She shares concrete builds: a folder-scanning script that has saved her team hours a week for over a year, and an investor sourcing pipeline cobbled together with Claude's MCP connectors across Apollo, HubSpot, Gmail, and more. The conversation is grounded by a distinction worth holding onto: automation a non-engineer can do tends to live in the task, not the full workflow. Ellen is candid about the limits, including a portfolio-monitoring tool she abandoned at two in the morning after overestimating what AI-assisted coding could do for her. Her larger point is that the constraint has moved. Coding is cheap now, so the scarce skill is judgment about what to build in the first place, and where buying still beats building. Key Takeaways * Non-engineers can reliably automate individual tasks, but full workflows still need real engineering. * AI is dependable on general financial analysis and weak on niche knowledge that lives outside its training data. * Knowing what to build next is now more valuable than the ability to build it. * Build versus buy still matters, because someone has to maintain, debug, and own what you build. * Start by automating one workflow you genuinely hate, then expand slowly. * Token maxing is a vanity metric; ROI and a clear North Star are the real measures. * Treat AI output like a strong team member's draft that you still review and stand behind. About the Podcast The A to Z of AI is hosted by Anthony Thomas and Zac Henderson and powered by The Suite. The show explores real-world AI use cases across industries, focusing on practical workflows, tools, and lessons learned from operators using AI in their daily work.

#15 - Tabs

Ali Hussain is co-founder of Tabs, a billing and revenue recognition platform built on an AI layer that reads contracts, usage data, and product context to automate downstream billing and rev rec workflows. In this episode, Ali walks through how Tabs was purpose-built around a problem that was fundamentally unsolvable before LLMs arrived: the commercial complexity between B2B companies is too variable, too contextual, and too fast-moving for traditional automation to handle. That premise now shapes both what Tabs builds and how the company itself operates internally. The broader conversation cuts through to something that often gets skipped in AI discussions: the cost layer. Ali makes the case that AI is already a utility, with real compute costs that most companies are still insulated from. As usage scales and vendors stop absorbing margin, that will change. Understanding AI as infrastructure, not just software, has direct implications for how operators budget for it, how they price their own products, and how they structure their teams going forward. Key Takeaways * AI is a utility, not a tool: compute costs are real, and CFOs will increasingly feel them as usage scales * The revenue function has shifted from rigid, annual pricing systems to continuous, strategic negotiation enabled by AI automation * The reporting and insight layer is where AI adds the most value in finance; deterministic, high-stakes execution still requires human traceability * Tabs embeds a "mini AI lead" within each functional team rather than centralizing ownership, keeping adoption close to the actual work * Ali runs without an EA: an AI agent handles scheduling, while Claude handles analysis and Gong handles call intelligence * Design and outbound workflows are seeing some of the largest time savings, with brand-trained agents replacing hours of manual work * Ali's one non-negotiable: first-pass writing stays human, because the thinking process behind it matters as much as the output itself About the Podcast The A to Z of AI is hosted by Anthony Thomas and Zac Henderson and powered by The Suite. The show explores real-world AI use cases across industries, focusing on practical workflows, tools, and lessons learned from operators using AI in their daily work.

#14 - Harvey

In this episode, Anthony Thomas and Zac Henderson are joined by Tara L. Waters, Legal Innovation Partner at Harvey. Drawing on her time as Chief Digital Officer at Ashurst, where she led one of the first firm-wide deployments of Harvey, Tara explains what changed when generative AI arrived: technology adoption in legal shifted from something operators had to push against resistance to something senior practitioners actively pulled toward. The conversation stays grounded in operational reality. Tara walks through how successful rollouts depend on meticulous planning, cross-functional teams, and treating adoption as a true transformation effort spanning governance, training, and performance expectations. She also offers a candid view on measuring success, the limits of efficiency metrics, and why she still feels early in her own AI journey despite being a recognized voice in the field. Key Takeaways * AI flipped legal technology adoption from a push by operators to a pull from senior practitioners * Successful rollouts depend on meticulous planning, cross-functional teams, and in-person activation events * A deliberate, phased approach now dominates over enterprise-wide deployments * Using AI as a thinking partner compresses days of mental processing into a single walk * Define value broadly, since how people feel about the work often matters more than time saved * The fastest way to start is to write down five tasks that annoy you and work through them with AI About the Podcast The A to Z of AI is hosted by Anthony Thomas and Zac Henderson and powered by The Suite. The show explores real-world AI use cases across industries, focusing on practical workflows, tools, and lessons learned from operators using AI in their daily work.

#13: Worksome

In this episode, Anthony Thomas and Zac Henderson are joined by Laura Jeffords Greenberg, General Counsel at Worksome. Laura brings a perspective that is rare in legal leadership: a builder's mindset shaped by nearly a decade in European tech, stints at Unity Technologies and Wordsmith, and a deliberate focus on using AI to do more with less. The conversation covers how she has structured her legal workflows around Claude's skills and Coworker platform, why AI caught things two humans missed in a settlement review, and how she thinks about the real constraints of using general-purpose AI tools in a regulated environment. What makes this episode distinct is Laura's refusal to treat AI as a productivity shortcut layered on top of existing habits. Her approach is structural: identify what needs to happen legally, build the skill or plugin to handle it, and reserve human judgment for the work that actually requires it. That reframing has implications well beyond legal teams, for any function trying to figure out where AI creates genuine leverage and where it still falls short. Key Takeaways: * AI can outperform two human reviewers on accuracy, not just speed: Laura's settlement review took seven minutes and surfaced errors neither she nor her colleague caught * Skills in Claude Coworker are essentially structured prompts; start by building one workflow that already lives in your head and give it to AI to execute * Wrapping individual skills into a plugin creates a reusable legal command center without requiring constant manual setup * General-purpose AI tools have real limits in regulated environments: audit trails, confidentiality, and privilege are legitimate constraints, not excuses to avoid adoption * Team structure should lead into individual strengths, not try to lift weaknesses; AI removes the work people hate, freeing up the work they're actually good at * Curiosity is the foundational skill; if you cannot use AI at work, use it personally until the instinct to reach for it becomes automatic * Agents are the near-term frontier: skills will evolve into agents that communicate with each other and resolve tasks with minimal human input About the Podcast The A to Z of AI is hosted by Anthony Thomas and Zac Henderson and powered by The Suite. The show explores real-world AI use cases across industries, focusing on practical workflows, tools, and lessons learned from operators using AI in their daily work.

#12: Lockton

Preet Gill, EVP at Lockton and head of the firm's global technology practice, joins Anthony Thomas and Zac Henderson to discuss how the world's largest privately held insurance brokerage is embedding AI across both client-facing and internal operations. The conversation covers Lockton's structured approach to AI adoption, from a formal governance group to day-to-day workflow automation, and how being independent and privately held gives them room to move faster than legacy competitors. The episode takes a sharp turn into territory rarely covered on the show: what AI means for insurability itself. Drawing on his work building coverage frameworks for autonomous vehicles and cyber risk, Preet makes the case that AI liability is on the same trajectory as cyber insurance and will become its own distinct coverage category, underwritten around performance rather than failure. Key Takeaways: * Lockton runs two parallel AI tracks: an external client advisory hub and an internal associate productivity hub * AI analyzing existing data and documents is more reliable than generative tasks; hallucinations drop when the model is given structured, task-specific data * A proposal and presentation builder tool is saving Lockton teams 1 to 2 hours per week, with formal measurement now in place * Team roles are shifting from producing raw content to interpreting outputs, challenging AI results, and applying judgment * Governance investment should come early, particularly in regulated industries; data usage policies and acceptable-use guidelines need to precede broad deployment * AI liability insurance will follow the same path as cyber: starting as a clause inside general liability, then becoming its own underwriting category as claims precedent builds * The key risk in algorithmic decision-making is underperformance, not outright failure, which requires fundamentally different coverage design About the Podcast The A to Z of AI is hosted by Anthony Thomas and Zac Henderson and powered by The Suite. The show explores real-world AI use cases across industries, focusing on practical workflows, tools, and lessons learned from operators using AI in their daily work.