Podcast
AI, Actually
Hosted by AnswerRocket · EN
Tired of the AI hype? So are we. Welcome to AI, Actually: the podcast that cuts through the noise and gets real about how artificial intelligence can work for your business. In each episode, our resident AI and business transformation experts–along with occasional industry guests–hold a candid, jargon-free conversation on what it takes to get actual value from AI.
Join us as we tackle topics like: the real difference between the latest LLM models, why generic AI can't make sense of your messy company data, how to get your GenAI use case off the ground, and what the rise of AI agents means for your business. This is your practical playbook for putting AI to work. No PhD required.
AI, Actually is produced by AnswerRocket. Since 2013, our enterprise AI solutions have helped Fortune 500 companies achieve measurable results through their AI transformations. This podcast is where we share what we’ve learned.
16episodes
Episodes
Newest firstAll episodes
Autonomous Agents in the Enterprise and How AI is Disrupting SaaS
Feb 1000:51:59Tap to summarizeArtificial intelligence has crossed a major threshold. AI is no longer confined to answering questions in a chat window. It's sending emails, scheduling meetings, moving data between systems, and operating with its own credentials and access. This shift from suggestion to action is forcing every enterprise leader to rethink their AI strategy.In this episode, Pete Reilly sits down with Shanti Greene, Jim Johnson, and Stew Chisam to examine three developments reshaping how businesses approach AI. From the OpenClaw experiment that briefly turned the internet into a Skynet-like scenario, to Anthropic's Cowork tool bringing agent capabilities to non-technical workers, to the fundamental question of whether enterprises should build their own software instead of buying it—this conversation tackles the practical implications of AI that actually does things.Topics covered:✅ How OpenClaw (formerly ClawdBot/MoltBot) represents the future of AI agents with their own identity, proactive behavior, and inter-agent communication✅ Why enterprise leaders need to start thinking about "onboarding non-human employees" with email addresses, Slack accounts, and human supervisors✅ Claude Cowork's approach to bringing agent capabilities to business users through local file access and workflow automation✅ The shifting build-vs-buy decision as AI makes custom software development dramatically faster and cheaper✅ Whether traditional SaaS companies like Salesforce can survive in an AI-first worldFollow the Gang:Stew Chisam, Operating Partner, StellarIQ - https://www.linkedin.com/in/stewart-chisam-7242543/ Jim Johnson, Managing Partner, AnswerRocket | https://www.linkedin.com/in/jim-johnson-bb82451/Pete Reilly, COO, AnswerRocket - https://www.linkedin.com/in/petereilly Shanti Greene, Head of Data Science and AI Innovation, AnswerRocket - https://www.linkedin.com/in/shantigreene/ Chapters:00:00 Introduction: The Evolution of AI in Enterprises01:20 OpenClaw: The Next Iteration of AI Assistant04:18 Implications of AI Agents in the Workplace10:01 Understanding the Difference: Traditional AI vs. OpenClaw14:48 Navigating Security and Accountability in AI21:29 The Future of Non-Human Employees in Enterprises26:42 Exploring Claude Cowork: A Shift in Focus39:29 The Build vs. Buy Debate in CRM Development#AIAgents #AutonomousAI #OpenClaw #ClaudeCowork #EnterpriseAIStrategy #AIWorkflows #AgenticOperations #BuildVsBuy #SaaSDisruption #AIEmployeeOnboarding
Transcribe →Building Software 10x Faster with AI: A Real-World Walkthrough
Jan 2700:45:15Tap to summarizeWhat happens when AI coding agents enter the development process? AnswerRocket's team built a production-ready CRM in just four weeks, showing what's possible with next-generation software development.In this episode, Pete Reilly, Alon Goren, Mike Finley, and Andy Sweet pull back the curtain on what modern AI development actually looks like in practice. Using a custom CRM project as their case study, they demonstrate how coding agents, contextual design, and new architectural patterns are fundamentally rewriting the build versus buy equation for enterprise software.This isn't theory—it's a live demonstration of software that would have taken 10x longer to build just six months ago, with insights on what this means for development teams, IT strategy, and the future of business software.Topics covered:How AI coding agents reduced development time by 10x for production-grade softwareWhy context is now more valuable than data entry in modern CRM designThe shift from "system of record" to "sales assistant" in CRM philosophyHow playbooks and workflows enable AI to follow company-specific processesWhy the build vs. buy decision is becoming "build with composable blocks"What "vibe coding" means and why offshore development strategies need rethinkingHow maintenance changes when AI agents can read source code directlyFollow the GangMike Finley, CTO, AnswerRocket - https://www.linkedin.com/in/mikefinley/ Pete Reilly, COO, AnswerRocket - https://www.linkedin.com/in/petereilly Andy Sweet, VP Enterprise AI Solutions, AnswerRocket - https://www.linkedin.com/in/andrewdsweet/ Alon Goren, AnswerRocket, CEO - https://www.linkedin.com/in/alon-goren-87889681/ Chapters00:00 Intro: Demo to AI in Software Development02:10 Understanding Customer Interactions and CRM Needs04:39 Reimagining CRMs in the Age of AI06:00 Demo Walkthrough of the CRM20:02 What's Possible Now with AI-Assisted Development21:25 Designing an AI-Compatible Stack27:32 Flipping the Build vs. Buy Dilemma32:59 AI's Impact on Offshore Development35:57 The Future of Business and Software Customization39:23 Maintaining AI-Assisted Software Solutions#AIDevelopment #CodingAgents #CRMCustomization #VibeCoding #BuildVsBuy #EnterpriseSoftware #AIAgents #SalesAutomation #SoftwareDevelopmentLifecycle #ClaudeCode #AgenticDevelopment #ContextDrivenDesign
Transcribe →AgentOps: Why Keeping AI Agents Running Is Harder Than Building Them
Jan 1300:39:15Tap to summarizeAs enterprises deploy AI agents into production, a new operational challenge emerges: how do you monitor and maintain systems that don't fail with error codes, but instead drift subtly away from expected performance? In this episode, the AI, Actually crew tackles the emerging discipline of AgentOps—the practice of keeping AI agents performing at peak business value over time.The discussion cuts through the hype around "self-learning" and "automated" AI to reveal the hard truth: agentic systems require continuous human oversight, just like human employees do. From probabilistic model behavior to reasoning model complexity, the team explores why traditional IT monitoring approaches fall short and why businesses need to rethink who owns these digital workers.Topics covered:Why AgentOps is fundamentally different from traditional DevOps and LiveOpsThe three levels of agent complexity and three types of drift that can derail performanceWhy traditional IT support models don't work for goal-driven agentic systemsThe organizational challenge of bringing together business knowledge, AI expertise, and technical skillsWhy there's no "blue screen of death" for agent failures—and what that means for monitoringFollow the Gang:Nicole Kosky, Senior Director of Services, AnswerRocket | https://www.linkedin.com/in/nicole-kosky-5b9a3b6/Joey Gaspierik, Director of Enterprise Sales, AnswerRocket | https://www.linkedin.com/in/joey-gaspierik-4a613642/Stew Chisam, Operating Partner, StellarIQ - https://www.linkedin.com/in/stewart-chisam-7242543/ Jim Johnson, Managing Partner, AnswerRocket | https://www.linkedin.com/in/jim-johnson-bb82451/Chapters:00:00 Introduction to AgentOps02:38 Defining Agentic Operations09:30 The Role of Human Oversight11:01 Understanding Performance Degradation17:27 The Complexity of Monitoring Agents26:07 Organizational Challenges in AgentOps31:00 The Future of Agentic Operations35:26 What's An Agent?Hashtags: #AIImplementation #EnterpriseAIAdoption #OrganizationalChange #AILiteracy #GeneralistEngineers #LastMileProblem #VibeCoding #AIROI #RevenueGeneration #OpenAIWhitePaperKeywords: Agent Ops, AI agents, enterprise AI, LLM monitoring, model drift, probabilistic systems, agentic AI, AI operations, AI governance, AI deployment, DevOps, LiveOps, reasoning models, AI scalability, digital workers
Transcribe →Open AI’s Playbook for Scaling AI, Why Generalists Are Winning, and Revenue-Driven ROI
Dec 1600:35:48Tap to summarizeThe old IT playbook is officially dead. Quarterly release cycles, endless approval committees, and throwing requirements over the wall? None of that works when AI models evolve every few weeks. In this episode, Pete Reilly sits down with Jim Johnson, Alon Goren, and Shanti Greene to unpack OpenAI's new white paper, "From Experiments to Deployments: A Practical Path to Scaling AI," and share what they're seeing on the ground with enterprise clients.This isn't theory—it's a frontline report from teams who are helping Fortune 500 companies navigate the messy reality of AI adoption. The conversation tackles the organizational upheaval required to move fast, the "last mile problem" that kills so many AI projects, and why the future belongs to curious generalists who can bridge business and technology. They also explore how AI is shifting ROI conversations from cost savings to revenue generation, and why vibe-coding a working prototype in hours is now entirely possible (with some important caveats about technical debt).Topics covered:Why IT and business teams must close the collaboration gap to succeed with AIThe "last mile problem": getting from 80% complete to actually usefulHow generalists with medium depth are becoming more valuable than deep specialistsBuilding an AI-literate workforce through curiosity and continuous learningShifting from cost-savings ROI to revenue-generating AI productsThe reality of vibe-coding: prototypes in hours, but engineering rigor still mattersFollow the Gang:Pete Reilly, COO, AnswerRocket - https://www.linkedin.com/in/petereilly Shanti Greene, Head of Data Science and AI Innovation, AnswerRocket - https://www.linkedin.com/in/shantigreene/ Alon Goren, Founder, AnswerRocket | https://www.linkedin.com/in/alon-goren-87889681/Jim Johnson, Managing Partner, AnswerRocket | https://www.linkedin.com/in/jim-johnson-bb82451/Chapters:00:00 Introduction to AI Deployment Challenges02:52 The Shift from IT to Business Collaboration05:41 The Role of Generalists in AI Integration08:42 Redefining AI Projects and Business Goals11:35 Accelerating Development Timelines with AI14:47 Building an AI Literate Workforce17:45 The Changing Landscape of ROI in AI20:58 Final Thoughts on Embracing AI#AIImplementation #EnterpriseAIAdoption #OrganizationalChange #AILiteracy #GeneralistEngineers #LastMileProblem #VibeCoding #AIROI #RevenueGeneration #OpenAIWhitePaperKeywords: AI implementation, enterprise AI adoption, organizational change, AI literacy, generalist engineers, last mile problem, vibe coding, AI ROI, revenue generation, OpenAI white paper
Transcribe →Google's Gemini 3 Breakthrough and Bold AI Predictions for 2026
Dec 200:36:05Tap to summarizeThe race for AI dominance just shifted. Google's Gemini 3 launch was a coordinated ecosystem play that could reshape how enterprises think about AI infrastructure. In this episode, Pete Reilly sits down with Andy Sweet, Shanti Greene, and Stew Chisam to dissect what Gemini 3 really means for enterprise adoption, where the technology is genuinely improving, and where marketing hype obscures practical limitations.The conversation moves beyond surface-level benchmarks to tackle the uncomfortable reality facing IT leaders: foundation models are converging on performance, but the real competitive advantage lies in how you architect solutions on top of them. The team explores Google's commanding lead in multimodal capabilities, the strategic implications of vendor ecosystems, and why enterprises betting everything on a single provider might be making a costly mistake. Then they close with their boldest predictions for 2026, from content exhaustion and the death of infographics to agents working autonomously longer than your employees.What You'll Learn:How Gemini 3's pre-training approach signals continued model improvements and what that means for the scaling law debateWhy Google's multimodal dominance (backed by YouTube, Google Photos, and Drive) creates a moat that competitors can't easily replicateThe critical difference between general intelligence benchmarks and enterprise intelligence that understands your business contextWhy the "stateless" problem keeps plaguing AI solutions and how memory scaffolding becomes essential for business applicationsPredictions for 2026: content exhaustion, shadow IT proliferation, and the moment enterprises realize there's no easy buttonFollow the Gang:Shanti Greene, Head of Data Science and AI Innovation, AnswerRocket - https://www.linkedin.com/in/shantigreene/ Pete Reilly, COO, AnswerRocket - https://www.linkedin.com/in/petereilly Andy Sweet, VP Enterprise AI Solutions, AnswerRocket - https://www.linkedin.com/in/andrewdsweet/ Stew Chisam, Operating Partner, StellarIQ - https://www.linkedin.com/in/stewart-chisam-7242543/ Chapters: 00:00 Introduction to Gemini 3 and Episode Overview 01:45 Gemini 3's Long-Term Planning Capabilities 04:17 Are LLMs Becoming Commoditized Primitives? 08:04 Model Specialization and Jagged Edges 11:15 Why Multi-Vendor Strategy Matters for Enterprises 13:29 OS/2 vs Windows: Best Doesn't Always Win 15:10 The Scaling Law Debate and Pre-Training Improvements 16:12 Enterprise Intelligence vs General AGI 22:41 How Enterprises Should Think About Gemini 3 26:44 Bold Predictions for 2026 28:08 Content Exhaustion and the Infographic Problem 33:06 Agents as Autonomous Team Members#Gemini3 #EnterpriseAIArchitecture #MultimodalAI #AIAgents #VendorLockIn #SemanticLayer #ScalingLaws #PreTrainingCompute #EnterpriseIntelligence #AIPredictions2026Keywords: Gemini 3, enterprise AI architecture, multimodal AI, AI agents, vendor lock-in, semantic layer, scaling laws, pre-training compute, enterprise intelligence, AI predictions 2026
Transcribe →Breaking Down Nate B. Jones' 6 Engineering Principles for AI Agents
Oct 7Tap to summarizeTired of AI agents that forget context mid-conversation or drift subtly off course in production? You're not alone. In this episode, the AI, Actually crew unpacks six critical engineering principles for building reliable AI agents—principles that separate proof-of-concepts from production-ready systems.Pete, Mike, Andy, and Stew break down insights from AI expert Nate B. Jones, translating technical concepts into business-focused guidance. They explore why AI memory isn't just about storage, how to bound uncertainty without killing creativity, and why monitoring AI systems requires a completely different approach than traditional software.This episode covers:Why stateful intelligence and memory management are fundamental to useful AI interactionsHow to engineer controls that bound uncertainty without over-constraining your modelsThe shift from binary failures to subtle quality drift in AI systemsCapability-based routing: matching the right model to the right jobPost-production monitoring strategies that catch problems before your users doContinuous validation techniques for multi-turn agent conversationsThis episode of AI, Actually centers around a video by @nate.b.jones about the 6 principles of AI Agents. That video can be watched in its entirety here: I've Built Over 100 AI Agents: Only 1% of Builders Know These 6 PrinciplesFollow the Gang:Mike Finley, CTO, AnswerRocket - https://www.linkedin.com/in/mikefinley/ Pete Reilly, COO, AnswerRocket - https://www.linkedin.com/in/petereilly Andy Sweet, VP Enterprise AI Solutions, AnswerRocket - https://www.linkedin.com/in/andrewdsweet/ Stew Chisam, Operating Partner, StellarIQ - https://www.linkedin.com/in/stewart-chisam-7242543/ Chapters: 00:00 Introduction to AI Agents and Engineering Principles01:34 Introducing Nate B. Jones' AI Engineering Principles03:03 Stateful Intelligence10:16 Bounded Uncertainty19:55 Intelligent Failure Detection20:51 Evaluating LLM Responses22:16 Monitoring Quality and Performance23:53 Active Maintenance of LLM Systems26:18 Understanding Subtle Failures26:55 Capability-Based Routing30:22 Aligning Models with Business Processes33:41 Nuanced Health State Monitoring37:36 Continuous Input Validation41:36 Closing ThoughtsKeywords: AI agents, agentic AI, AI engineering, AI memory, stateful intelligence, AI monitoring, capability-based routing, AI evaluation, production AI, enterprise AI, AI agent development, LLM engineering, AI testing, AI agent failures, AI system monitoring
Transcribe →