Ever pull up a trace, stare at a wall of spans, and just paste the whole thing into an LLM and ask "what happened here?" Feel seen, you're not alone. Traces are built for machines, not humans, which means every time something goes wrong, you're stuck reverse-engineering your own agent's decisions. Rasa Inspector skips that step entirely. Instead of raw trace data you have to decode (or outsource to another model just to make sense of), Inspector shows you the actual decision path live: which part of your agent activated, what memory items were filled, what's sitting on the stack, all human-readable, all in real time as you chat. No trace IDs. No copy-pasting into a chatbot to translate your own system back to you. Just the decisioning, laid out clearly. 👉 Docs: https://rasa.com/docs/pro/testing/trying-assistant/ 👉 Talk to a Rasa Product Expert: https://hellorasa.info/4hM5htv #RasaInspector #ConversationalAI #AIAgents #Debugging #Rasa

🤖 AI doesn't need to be *smarter.* It needs to know **you** better. Brian Alvey, CTO of WordPress VIP, believes the next breakthrough in AI won't be answering more questions—it's understanding your context. Instead of simply telling you the weather, imagine an AI that knows where you're headed, predicts what you'll need, and helps you make better decisions before you even ask. 🌤️ "It's 80° here..." ❄️ "...but you're driving into San Francisco, where it's 25 degrees colder. Grab a jacket." That's the kind of AI people will actually pay for. The future of enterprise AI isn't just better models—it's more personalized, context-aware experiences. 🎧 Watch the full conversation with Lauren Goerz and Brian Alvey on *The Dialogue Architects*. #ArtificialIntelligence #EnterpriseAI #ConversationalAI #WordPressVIP #AIProducts

AI isn't replacing websites—it's changing how people discover them. In this episode of The Dialogue Architects, Lauren Goerz sits down with Brian Alvey, CTO of WordPress VIP, to explore how AI answer engines, enterprise content management, and conversational experiences are reshaping the future of the web. Here's What You'll Hear Why WordPress still powers more than 40% of the web How enterprise CMS platforms differ from consumer websites Why AI answer engines are changing content discovery The new role websites play as systems of record for AI How organizations should govern chatbots and AI assistants Why evaluations and guardrails matter more than model choice The risks of autonomous agents communicating with other agents Applying Jobs to Be Done thinking to AI workflows Why accessibility should shape every conversational experience How enterprises can prepare for an AI-first future without losing control #ConversationalAI #WordPress #EnterpriseAI #AI #CMS #WordPressVIP #CustomerExperience #TheDialogueArchitects

🤖 Something unexpected happened after ChatGPT. People stopped typing keywords... ...and started having conversations. Aleks Bass explains how AI has fundamentally changed user behavior—from quick commands to detailed stories, life experiences, and long-form responses. That's creating incredible opportunities... ...and brand new challenges for every AI product team. 🎧 Full episode of *The Dialogue Architects* available now. #ConversationalAI #ChatGPT #AI #Typeform #UserResearch

Learn how a Rasa agent handles a basic FAQ conversation: project structure, the Command Generator, dialogue understanding, and dialogue management, all in one chapter. This is Chapter 1 of Rasa University, the free developer certification path from first response to production agents. You'll make changes to a working project, train, and test your agent. By the end you'll know what each file does and what actually happens on every turn of a conversation. WHAT THIS CHAPTER COVERS ▸ The Rasa project structure: what each file does and why ▸ How the Command Generator turns user input into structured next steps ▸ Dialogue understanding vs. dialogue management: two jobs, one system ▸ Basics of building flows ▸ Contextual Response Rephraser ▸ Conversation patterns: the moments most agents fail at ▸ Training your agent and testing it after every change ▸ Building your first flow START THE COURSE ▸ Rasa University (free, self-paced, certificate on completion): https://hellorasa.info/4h2NHB8 ▸ Free developer license: https://hellorasa.info/44OASTI ▸ Community Discord: https://hellorasa.info/4f4cqSV ▸ Documentation: https://rasa.com/docs Rasa is the developer platform for enterprise AI agents. It includes an open-code agentic AI Framework that runs in your infrastructure. Each chapter builds on the last. Subscribe for the rest of the series. #Rasa #AIAgents #AgenticAI #Python #DeveloperTutorial

🤖 Everyone asks whether AI will replace researchers. Aleks Bass sees a different future. Instead of spending hours manually coding every interview and comment, AI can handle the repetitive work—giving researchers more time to validate insights, spot patterns, and ask better questions. ✨ More context. 📊 Better analysis. 🧠 Smarter decisions. The future of research isn't AI *instead* of humans. It's AI helping humans understand people better. 🎧 Watch the full conversation with Aleks Bass, CPTO of Typeform, on *The Dialogue Architects*. #AI #CustomerResearch #QualitativeResearch #ProductManagement #Typeform

Install Rasa and get your first AI agent running in Rasa University. This is a tutorial walking you through the labs in the prerequisites chapters: the complete developer certification pathway from setup to your first Rasa Agent. Sign up for the course and get certified on Rasa here: https://hellorasa.info/4h2NHB8 BEFORE YOU START, YOU'LL NEED: ▸ An OpenAI API key ▸ A Rasa developer license: https://hellorasa.info/44OASTI ▸ Basic familiarity with the command line WHAT THIS CHAPTER COVERS: ▸ Installing the Rasa Framework ▸ Setting up your Python environment ▸ Verifying your installation ▸ What you'll build in the chapters ahead CONTINUE LEARNING: 🎓 Rasa University: https://hellorasa.info/4h2NHB8 📄 Documentation: https://rasa.com/docs 💬 Community Discord: https://hellorasa.info/4f4cqSV ⭐ GitHub: https://github.com/RasaHQ Rasa is the developer platform for enterprise AI agents. The Framework is code-first and runs in your infrastructure: on-premise, private cloud, or air-gapped. Subscribe for the rest of the series. Each chapter builds on the last, ending with a production-ready agent. #Rasa #AIAgents #Python #AgenticAI #DeveloperTutorial

In this episode of The Dialogue Architects, host Lauren Goerz sits down with Aleks Bass, Chief Product & Technology Officer at Typeform, to explore why customer research is becoming even more important in the age of AI. Here's What You'll Hear Why customer research becomes more valuable in the age of AI How Aleks' psychology background influences product strategy The three principles Typeform uses to build better products Why shared language creates better collaboration across teams How AI-powered data enrichment improves customer experiences Inside Research Flow and AI-moderated customer interviews Why qualitative research still matters at enterprise scale How structured processes accelerate AI product development Why customer feedback beats AI simulation every time The future of AI-first product organizations Timestamps (00:00) Meet Aleks Bass (00:44) From Psychology to Product Leadership (01:22) Three Principles for Building Better Products (04:53) Typeform's Evolution Beyond Forms (07:10) What It Means to Be AI-First (08:29) AI-Powered Data Enrichment (11:18) Designing Better Customer Conversations (14:25) Inside Research Flow (17:42) Why Product Fundamentals Still Matter (20:02) Scaling Customer Research with AI (26:08) Managing AI Product Teams (30:24) Why Process Is Your AI Superpower (36:24) Customer Conversations vs AI Simulation (38:49) Closing Thoughts About the Guest Aleks Bass is the Chief Product & Technology Officer at Typeform, where she leads product strategy, engineering, and AI innovation. Drawing on her background in social psychology and market research, Aleks champions a research-first approach to product development that prioritizes understanding customer behavior before building solutions. Today, she is leading Typeform's transformation into an AI-first customer intelligence platform, helping organizations collect richer insights, automate research, and make better product decisions through a combination of human-centered design and artificial intelligence.

Rasa University is live. This video shows you where to find it, how to sign up, and how to start the Rasa Developer Certification (Foundations) course. ▶ Start the course: https://hellorasa.info/4h2NHB8 The Developer Foundations Certification course is free and self-paced, and takes around 3.5 hours. You'll build a working agent chapter by chapter, from your first responses to autonomous sub-agents with tool-calling, and earn your Rasa Developer Certificate when you complete all six chapters. What you need to start: - A free Rasa Developer Edition license: https://hellorasa.info/44OASTI - An OpenAI API key - Comfort with the command line and Python Questions while you're building? Find us in #rasa-university on Discord: https://hellorasa.info/4f4cqSV

What does success with AI customer support really look like? In this clip from The Dialogue Architects, Dino Letic of JetBrains explains why the biggest win isn't CSAT scores, deflection rates, or analytics—it's seeing a customer get the help they need instantly. For a longtime support engineer, the real dopamine hit comes when an AI assistant solves a problem the same way a human expert would. #AI #CustomerSupport #JetBrains #ArtificialIntelligence #CustomerExperience #CustomerService #TechLeadership #SupportEngineering #Automation #TheDialogueArchitects

In this episode of The Dialogue Architects, host Lauren Goerz sits down with Dino Letic, Product Manager at JetBrains, to explore what it takes to build conversational AI for developers. Unlike most customer support audiences, developers arrive with detailed logs, highly technical questions, and unique environments that make support particularly challenging. Dino shares how his background in support engineering shaped his approach to building JetBrains' unified support bot—an AI-powered experience designed to simplify customer interactions across a growing portfolio of products. The conversation explores the challenges of fragmented support systems, the importance of creating a single point of entry for users, and how the bot identifies intent, asks follow-up questions, summarizes issues, and escalates conversations when needed. Dino explains how JetBrains balances automation with human expertise, why knowledge management is critical to AI success, and how the team handles the ongoing challenge of outdated documentation and rapidly evolving products. Lauren and Dino also discuss measuring success, team ownership models, deployment considerations, and why JetBrains chose an on-premises AI approach. The episode concludes with a thoughtful discussion about the opportunities—and risks—of agentic AI in customer support and product decision-making. Whether you're building AI assistants, managing support operations, or designing enterprise conversational experiences, this episode offers practical lessons from one of the most technically demanding support environments in software. Here's What You'll Hear Why developers are uniquely challenging support customers How JetBrains unified support across multiple products and teams The role of AI in ticket triage, escalation, and issue summarization Why a single conversational entry point improves customer experience How JetBrains manages knowledge sources and documentation quality The challenges of outdated information in AI-powered support What success metrics matter most for support automation Why JetBrains chose an on-prem deployment strategy Team structures, ownership models, and operational lessons Where agentic AI fits—and where caution is required #ConversationalAI #CustomerSupport #DeveloperExperience #AIAgents #EnterpriseAI #JetBrains #Rasa

Everyone is talking about prompt engineering. But Rebecca Evanhoe thinks the real advantage is somewhere else: 👉 Data. As AI teams move away from traditional scripting, they're spending less time tweaking individual words and more time evaluating outcomes at scale. The challenge? You can't evaluate AI without quality data. The future isn't prompt-first. It's evaluation-first. And evaluation runs on data. #AI #LLM #PromptEngineering #EnterpriseAI #DataScience

Ever notice how some AI voice assistants feel unnatural? It's often not what they're saying—it's when they're saying it. Rebecca Evanhoe of Slang AI explains that natural conversation depends on timing, pacing, and smooth turn-taking. When an LLM takes too long to respond, the entire interaction starts to feel robotic. That's why voice AI teams obsess over latency, model size, and response timing—not just intelligence. The future of voice AI isn't just smarter responses. It's more natural conversations. #VoiceAI #ConversationalAI #LLM #ArtificialIntelligence #SlangAI

🎙️ Most teams think the secret to better AI is better prompts. Rebecca Evanhoe says that's the wrong question. At Slang AI, handling millions of restaurant phone calls every month means success isn't measured by clever prompts—it's measured by whether the system actually works. The real differentiator? ✅ Rigorous evaluation ✅ Transcript reviews ✅ Human scoring ✅ Automated testing ✅ Continuous improvement As Rebecca tells Lauren Goerz on The Dialogue Architects: "The future belongs to teams that can evaluate AI, not just build it." Watch the full episode to learn how enterprise teams are creating AI experiences users can trust. #ConversationDesign #VoiceAI #EnterpriseAI #LLMEvaluation #SlangAI (00:00) Introduction to The Dialogue Architects (00:39) Meet Rebecca Evanhoe of Slang AI (01:30) A Day in Product Management & Voice AI (02:25) Why Conversation Design Still Matters (03:29) Why LLM Evaluation Is Critical (05:38) Is Conversation Design Dead? (06:19) Prompt Engineering vs Conversation Design (07:07) What Conversation Designers Actually Do (10:01) User Advocacy, Research & Design Workshops (12:22) Hybrid AI vs Fully Generative Systems (16:31) When Deterministic Systems Beat LLMs (19:31) AI Costs, Latency & Model Selection (21:07) How Slang AI Evaluates LLM Performance (25:41) Transcript Reviews That Improve AI (26:34) Finding Root Causes Through Metrics (27:54) CSAT vs Sentiment: What Matters More? (31:07) Choosing the Right North Star Metrics (31:59) The Biggest Challenges in Voice AI Design (36:23) Multilingual AI & The Future Product Roadmap (40:31) Turning Customer Calls Into Business Insights (43:12) How AI Is Changing Human Communication (46:46) Rebecca Evanhoe's Advice for Conversation Designers (47:51) Closing Thoughts & Where to Connect

In this clip from The Dialogue Architects, Eugenia Zeibig, a marketing AI leader in pharma, shares why enterprise AI only becomes useful when it understands the nuance of the business it serves — and why trustworthy outputs matter more than abundant ones. Watch the full episode: From Reports to Responses: AI in Pharma Marketing Views shared are Eugenia Zeibig's own. #EnterpriseAI #ConversationalAI #PharmaMarketing #AIInnovation #MarketingAI #HealthcareAI #Rasa

In this clip from The Dialogue Architects, Lauren Goerz of Rasa explores why enterprise AI only works when the system understands the specific questions a business needs answered. Natural language is quickly becoming the interface for data analysis — and conversational AI is shaping that shift. Watch the full episode: From Reports to Responses: AI in Pharma Marketing #agenticai #ConversationalAI #EnterpriseAI #DataAnalytics #MarketingAI #HealthcareAI #Rasa

Enterprise AI isn't just about automation — it's about helping people make better decisions, faster. In this episode of The Dialogue Architects, host Lauren Goerz sits down with Eugenia Zeibig, Strategy & Innovation Lead — Marketing AI & Commercial Analytics at Pfizer, to explore how AI is reshaping decision-making in pharma marketing. The conversation focuses on what it takes to bring AI into the daily workflow of pharma marketers — combining visual insights with natural-language dialogue to move from weeks of stitched-together analysis to faster, more informed strategic decisions. Lauren and Eugenia discuss the challenge of designing AI systems that marketers can actually trust, why quality matters more than quantity when delivering recommendations, and how AI behavior needs to adapt across different therapeutic categories in pharma. The episode also explores: Balancing "push" insights with "pull" conversational querying Teaching AI to "speak marketing" The difference between high-impact strategic decisions and rapid campaign optimization loops Measuring AI adoption inside a marketing organization The future of orchestration across copilots, foundation models, and enterprise AI tools A practical conversation for enterprise AI leaders, marketers, product teams, and conversational AI practitioners interested in how natural language interfaces are reshaping data analysis and strategic decision-making in regulated industries. Views shared by Eugenia Zeibig are her own and do not reflect Pfizer corporate guidance. Guest Bio Eugenia Zeibig is the Strategy & Innovation Lead — Marketing AI & Commercial Analytics at Pfizer. With nearly two decades of experience connecting analytics to commercial outcomes across healthcare and retail, Eugenia leads initiatives focused on building AI-forward, data-driven solutions that empower enterprise marketing teams with conversational AI, strategic insights, and faster decision-making workflows. Key Topics Discussed Approaching AI design for pharma marketing teams Why analysis speed matters in pharma marketing Combining visual insights with natural-language dialogue Designing AI systems marketers can trust Why "quality over quantity" matters in AI recommendations Why AI behavior differs across therapeutic categories Strategic decision loops vs rapid campaign optimization How to measure AI adoption inside a marketing organization Empowering marketers through conversational analytics The future of orchestration across enterprise AI systems Episode Timestamps (00:00) Welcome and guest introduction (01:01) Eugenia Zeibig's role at Pfizer (02:50) Understanding pharma marketing audiences (04:55) The marketer workflow before AI (08:00) Designing AI for pharma marketers (10:24) Balancing push and pull interactions (13:07) Building trust through quality (15:23) Big-loop vs small-loop decision-making (18:46) Scaling AI across brands and therapeutic areas (22:59) Measuring adoption inside a marketing organization (27:07) Natural-language dialogue for marketers (29:24) Empowering marketers and data science teams (31:22) The future of AI tools and orchestration (33:51) Closing thoughts and farewell About The Dialogue Architects The Dialogue Architects is a podcast from Rasa, hosted by Lauren Goerz, exploring the craft and strategy behind designing conversational AI in the enterprise. Each episode brings together technologists, designers, and product leaders to unpack how dialogue is built, scaled, and governed in real-world systems.

We loved listening to Sapana Patel, CIO at Spirit Airlines, talk about her team and what actually motivates people in tech. She describes technologists as builders at heart. People who like to design, experiment, and then see their work come to life in real systems. Motivation isn’t abstract: It’s seeing products work, help users, and create real impact, whether that’s growth or efficiency for the business.

There are loads of CX metrics you can capture in the context of AI-driven customer conversations, but the one that has the highest signal-to-noise ratio is solution ratio. Here we talk through the pros and cons of that metric.

We’ve never had this level of user context before. Soon, brands will be able to personalize every interaction in real time by injecting the right data at the right moment into AI systems. This is what the future of customer experience looks like: context-aware, dynamic, LLM-driven interactions. Insights from Etie Hertz (GM of Conversational Intelligence @Contentsquare) #conversationalai #LLM #customerexperience #personalization #AI

Customer support isn’t a cost center; it’s a growth lever. Great support interactions increase LTV, retention, and customer spend. But most companies don’t measure it properly. Insights from Etie Hertz (former CEO of Loris.AI, now GM of Conversational Intelligence at Contentsquare). #customersupport #LTV #retention #customerexperience

We set up an agent harness — Claude Code, a Rasa MCP server, and some markdown files — and use it to build a flight upgrade assistant from scratch, live on camera. The airline bot on WhatsApp is actually pretty slick. You're already logged in, your bookings are there. But the conversation feels like filling in a form. Say anything off-script and it's lost. So we build something better: an upgrade agent that can check availability, show seats, quote prices, and handle hesitation naturally. "That's a lot" doesn't break it — it re-asks. The whole thing is built in one sitting using Claude Code with Rasa's MCP tools. No slides, no prep. *Try it yourself:* https://gist.github.com/amn41/537a607ccbc3ca557e7f23a97233f0be *What's in the harness:* - Claude Code (the base) - Rasa MCP tools (train, test, inspect — without leaving the editor) - CLAUDE.md (workflow instructions that make Claude Code self-validating) *Links:* - *Rasa Docs:* https://rasa.com/docs - *Code from this video:* https://gist.github.com/amn41/537a607ccbc3ca557e7f23a97233f0be

Spirit Airlines CIO Sapana Patel joins host Lauren Goerz to break down how enterprise AI actually gets built — from data architecture and conversational agents to leading teams through rapid technological change. Sapana walks through Spirit's four-layer AI framework, how her team handles anonymous customer data without sacrificing security, and what metrics truly define AI success at scale: solution ratio, revenue impact, uptime, and stability. She also shares her philosophy on technology leadership — treating engineers as artists, balancing innovation with operational excellence, and building trust across technical and business teams. If you work in AI, enterprise tech, or leadership, this conversation delivers the specifics most people skip over. 00:00 Introduction 02:30 Sapana Patel's approach to technology leadership 06:15 Spirit's four-layer AI architecture explained 12:40 Building conversational agents in-house 18:55 Handling anonymous customer data securely 24:10 Metrics that define AI success 30:45 Event-driven data platforms and customer touch points 38:20 Leading teams through AI transformation 44:00 Closing thoughts and leadership insights 🎙️ The Dialogue Architects drops new episodes regularly — subscribe and hit the bell so you never miss one. #TheDialogueArchitects #AILeadership #ConversationalAI #SpiritAirlines #CIO #TechLeadership #EnterpriseAI #DataStrategy #AIInfrastructure #CustomerExperience #ArtificialIntelligence #Podcast #TechPodcast #AITransformation #MachineLearning #DigitalTransformation #AviationTech #ProductivityAI #FutureOfWork #WomenInTech

This episode is a REPLAY of episode 1 featuring Rasa Founder Alan Nichol. Enjoy! We'll be back next week with new content.

In this episode of The Dialogue Architects, Lauren Goerz speaks with Etie Hertz, GM of Conversation Intelligence at ContentSquare, about the future of customer support when conversation quality, sentiment, and behavioral analytics come together. You’ll learn: • Why CSAT is fundamentally flawed • How Loris’s sentiment scoring (-2 to +2) works • What Conversational Quality (CQ) measures • How sentiment arcs can predict churn • Why tone, mirroring & empathy drive outcomes • Channel shifts: voice vs. digital support • Where friction hides in refund and policy flows • Cultural nuance in human and AI conversations • The role of AI agents—and why they raise new concerns Links & Resources Etie Hertz – https://www.linkedin.com/in/etiehertz/ Lauren Goerz – https://www.linkedin.com/in/lauren-goerz/ 🔗 Learn more about trustworthy conversational AI: rasa.com 🔗 Try Rasa today in the Rasa Playground: https://hellorasa.info/46UKlub Chapters included.

In this episode of The Dialogue Architects, Lauren Goerz sits down with Nick Thomson, General Manager of Got Bot, to explore the practical architecture behind high‑performing, enterprise‑grade messaging systems. Nick breaks down: • Why WhatsApp dominates customer interactions in South Africa • Workflow design vs. generative AI hype • Scaling commerce bots during traffic spikes • Integrations, failover plans, and operational risk • Why diagramming is essential for cross‑team alignment • MCP, agent orchestration, and multi‑bot ecosystems • Predictions for 2026: security failures, ROI pressure, agent skepticism ⏱ Chapters included in video description. Links & Resources Nick Thomson – https://www.linkedin.com/in/nick-thomson-51127a83/ Lauren Goerz – https://www.linkedin.com/in/lauren-goerz/ 🔗 Learn more about trustworthy conversational AI: rasa.com 🔗 Try Rasa today in the Rasa Playground: https://hellorasa.info/46UKlub (00:00) Welcome and guest intro (01:08) Nick’s origin story (05:32) Role today and GenAI reality check (06:59) WhatsApp commerce and scaling (10:31) Project priorities and workflow design (13:49) Diagramming with Excalidraw and Miro (18:14) MCP and agent orchestration (25:21) Customer war stories and ops challenges (29:42) Metrics, feedback, and containment (31:24) 2026 predictions and wrap‑up