Regular Expressions

Making AI-First Engineering Teams Successful, With Andy LaMora

Long before generative AI agents were writing code, platforms like Topcoder were using crowdsourcing to solve massive technical problems. It turns out that breaking down complex projects for a distributed network of human developers was the perfect training ground for how we interact with AI today. In this episode, Ben Griswold (Grizen) and Noah Heldman (OutcomeSource) welcome technology executive Andy LaMora to discuss the fascinating overlap between crowdsourced engineering and modern AI tools. Andy shares his insights on applying Daniel Kahneman's concepts of "fast and slow thinking" to software development, explaining why AI handles the fast, repeatable expertise while humans must retain the slow, deliberative problem-solving. The conversation explores how AI is actively changing the Software Development Life Cycle, the growing importance of building solid component catalogs, and a staggering real-world example of using Claude to build a fully compliant AWS banking infrastructure in just one month. Andy also shares his take on the Cursor versus Claude Code debate and answers the standard Five Questions, including a great story about stepping off the corporate ladder to ride a motorcycle across the country. In This Episode, You'll Learn: * How early crowdsourcing models set the blueprint for working with modern AI agents. * The application of "fast and slow thinking" to prompt engineering and system design. * Why the classic software development life cycle is evolving to favor small product engineering teams. * The story of building a fully functioning, regulated bank infrastructure on AWS in 30 days using spec-based coding. * A practical comparison of Cursor and Claude Code for different types of consulting and development workflows. * The Five Questions: Why we need to ban the word "paradigm," the undeniable productivity boost of walking your dog, and the value of taking an unconventional career break. Connect with Us: * Ben Griswold | Grizen: https://grizen.com [https://grizen.com] * Noah Heldman | OutcomeSource: https://outcomesource.com [https://outcomesource.com] * Andy LaMora on LinkedIn: https://www.linkedin.com/in/andylamora [https://www.linkedin.com/in/andylamora/]

AI Coding Is More About The Experience Than The Model

The tools we use to write software are evolving at breakneck speed. Cursor recently dropped version 3 featuring a brand new agent-first interface, while Anthropic continues to push developers toward the terminal with Claude Code. But are these new autonomous workflows actually an improvement for experienced engineers? In this episode, Ben Griswold (Grizen) and Noah Heldman (OutcomeSource) compare their hands-on experiences with the latest AI coding environments. They discuss the jarring feeling of having code abstracted away by terminal-based agents and the sheer frustration of hitting Claude Pro's usage limits after just ten minutes of work. The conversation also uncovers what a recent leak revealed about how Claude Code operates under the hood, proving that many of these advanced tools rely heavily on massive, hidden system prompts. In This Episode, You'll Learn: * The major changes in Cursor v3 and its new agent-first window. * Why terminal-based AI tools can disrupt a developer's natural workflow and visibility. * The reality of burning through tokens and hitting hard limits on expensive paid AI tiers. * What the Claude Code source leak showed us about system prompts and goofy loading verbs. * Why the industry is shifting from AI assistants to AI agents. * The reason Ben is actively searching for a Claude Code expert to prove his current workflow wrong. Connect with Us: * Ben Griswold | Grizen: https://grizen.com [https://grizen.com] * Noah Heldman | OutcomeSource: https://outcomesource.com [https://outcomesource.com]

Turning Top Performers Into AI Champions With Boz Vitanova

Rolling out AI tools in an enterprise requires a different strategy than a traditional software implementation. Because large language models are probabilistic and constantly evolving, top-down mandates often lead to poor adoption or employees quickly dismissing the technology as unreliable. In this episode, Ben Griswold (Grizen) and Noah Heldman (OutcomeSource) welcome Boz Vitanova, CEO of TeamLift. Boz explains how she helps companies identify high-performing, process-oriented employees and trains them to become internal "AI Champions." These champions discover best practices and build custom workflows from the bottom up, ensuring the technology solves actual business problems. The conversation also covers the most accurate ways to measure AI productivity, how AI tools are compressing the traditional product development lifecycle, and the unique challenges junior professionals face as they enter the workforce today. Plus, Boz takes on the Five Questions segment and shares a valuable startup lesson about trusting your own instincts over investor feedback. In This Episode, You'll Learn: * Why traditional software rollout strategies struggle to work for AI tools. * The specific traits to look for when identifying potential AI Champions in your organization. * Why measuring AI success by "time saved" or prompt volume can be misleading. * How tying AI workflows directly to existing company OKRs creates a better measure of success. * The ways AI is removing the "middle layers" of translation in product development. * The Five Questions: Why we need to retire the phrase "AI-powered" and the importance of stepping away from the screen for a walk. Connect with Us: * Ben Griswold | Grizen: https://grizen.com [https://grizen.com] * Noah Heldman | OutcomeSource: https://outcomesource.com [https://outcomesource.com] * Boz Vitanova | TeamLift: https://teamlift.co [https://teamlift.co]

The Bar Is Low In Tech Consulting. Do The Basics. Stand Out.

Ever feel like basic professionalism (like returning a phone call, showing up on time, or actually listening to a client) has become a rare superpower? You aren't alone. In fact, it might be the easiest competitive advantage you can build. In this episode, Ben Griswold (Grizen) and Noah Heldman (OutcomeSource) unpack a simple but profound truth about the modern workplace: the bar for success is shockingly low. They swap hilarious and frustrating customer service war stories, from Noah's 90-minute screaming match with a Vonage retention script to the dark magic of Dish Network's un-cancelable subscriptions. They contrast these nightmares with frictionless, positive experiences and apply the exact same lessons to technology consulting. From doing five minutes of basic AI research before a meeting to simply being a concise and patient listener, Ben and Noah explain how mastering the "controllables" can make you look like an absolute rockstar in any industry. In This Episode, You'll Learn: * Why the "bar is low" in modern business and how to use it to your advantage to win clients. * Hilarious customer service fails: The Vonage nightmare and the Dish Network retention trap. * A massive win: How Vercel uses AI chat to make cancellations and refunds completely frictionless. * The simple consulting pillars that win trust: Responsiveness, active listening, and detailed follow-through. * How to use AI to do highly effective client meeting prep in under 60 seconds. * Why controlling the "controllables" (punctuality, patience, conciseness) is the ultimate career hack for early-career engineers and seasoned veterans alike. Connect with Us: * Ben Griswold | Grizen: https://grizen.com [https://grizen.com] * Noah Heldman | OutcomeSource: https://outcomesource.com [https://outcomesource.com]

Rebecca George on AI, Bias, and Change Leadership

Organizations everywhere are rushing to implement AI, often measuring success by how many employees are actively using the tools. But treating AI adoption as a simple numbers game completely misses the mark. When speed becomes the primary goal, critical thinking is the first casualty. In this episode, Ben Griswold (Grizen) and Noah Heldman (OutcomeSource) welcome Rebecca George, an expert in organizational change management, AI strategy, and executive development. Rebecca shares a cautionary tale of using AI to synthesize interviews to save eight hours of work, only to spend two weeks cleaning up a political mess caused by algorithmic bias. The conversation explores why traditional change management no longer works, how Employee Resource Groups are the secret weapon for testing AI guardrails, and why parents and individuals must take AI safety into their own hands rather than trusting big tech. Rebecca also pulls from her background in theater to explain how leaders, particularly women in tech, can turn their professional triggers into their greatest superpowers. In This Episode, You'll Learn: * Why measuring AI success by license usage is a massive mistake. * The difference between outdated change management and true change leadership. * A real-world example of how AI bias created a stakeholder nightmare. * Why speed is a false currency when it comes to generative AI output. * How to use Employee Resource Groups to rigorously test AI platforms. * How principles from the theater can help you own your space in the boardroom. * The Five Questions: A plea to ban the phrase "human in the loop" and a hilarious story involving a major wardrobe malfunction in front of a college class. Connect with Us: * Ben Griswold | Grizen: https://grizen.com [https://grizen.com] * Noah Heldman | OutcomeSource: https://outcomesource.com [https://outcomesource.com] * Rebecca George | https://takeyourspace.com