TAKEAWAY - Data Products Done Right: Lessons from Capital One's Journey at Scale with Bethany Sehon

Shachar MEIR has spent 20 years fixing data teams and understanding why they fail. In this live episode, Shachar joins Juan and Tim to discuss the reason why they fail and what they should be doing in order to succeed.  Shachar argues that data leaders default to what they can control (cough cough technology_ and underinvest in people, process, and culture. A new CDO comes in, migrates legacy databases to a cloud data platform, and two years later you have the same problems on a better platform. We get into what's actually missing: building trust in data (which takes time and breaks fast), creating incentives so people actually use what data teams build, and spending real time in the business which is not asking what dashboards they need, but what problems they have.  Topics discussed:  Technology is not why data teams fail Why data leaders default to technology People, process and CULTURE Data trust Incentives Understanding the business Dashboard bloat Self-service analytics Dashboards aren't dead What CDOs do well and what they miss See omnystudio.com/listener for privacy information.

This is the takeaway episode with Shachar Meir, who has spent 20 years fixing data teams and understanding why they fail. In this live episode, Shachar joins Juan and Tim to discuss the reason why they fail and what they should be doing in order to succeed. See omnystudio.com/listener for privacy information.

Patrick McGarry, Federal CDO at ServiceNow and author of The Adaptive Organization, has spent years watching organizations accumulate data and tools but fail to convert them into decisions. Overhead grows. Outcomes don't. Hence “Intelligence Without Action Is Just Overhead.” Patrick joins Juan and Tim to walk through why federal data programs stall (the idea-to-deployment lag is long enough that the tech moves on), why incentives matter more than mandates, and why the right question should not be "how do we get AI ready" and instead should be "how do we make our organization accountable for AI."  Topics discussed:  Intelligence vs. action Ownership and outcomes Federal government's speed problem The CATALOG framework: Culture & Talent, Analytics & AI, Technology & Architecture, Alignment, Leadership & Governance, Operations & Delivery, Growth & Measurement. Governance as clarity, not bureaucracy Open standards, vendor lock-in, semantics, interoperability  Regulatory standards as a double-edged sword See omnystudio.com/listener for privacy information.

This is the takeaway episode with Patrick McGarry, Federal CDO at ServiceNow and author of The Adaptive Organization, who has spent years watching organizations accumulate data and tools but fail to convert them into decisions. Overhead grows. Outcomes don't. Hence “Intelligence Without Action Is Just Overhead.” See omnystudio.com/listener for privacy information.

Most companies slapping the "data product" label on existing data assets are fooling themselves. Bethany Sehon, Enterprise Data Leader at Capital One, has been building data products and ontologies in production long before they became buzzwords. Bethany joins Juan and Tim to share lessons learned: dedicated data product managers (not side-of-desk), ontologists treated as first-class citizens, a VP-accountable governance process, and much more.  Topics discussed:  Data products as a discipline, not a label Governance process as the forcing function  Dedicated roles, not side-of-desk work  Ontology as a strategic function  Standardization is necessary but not sufficient Less is more on scope  Trust is consumer-specific AI expands what counts as a data product See omnystudio.com/listener for privacy information.

This is the takeaway episode with Bethany Sehon, Enterprise Data Leader at Capital One where she share lessons learned of building data products and ontologies in production long before they became buzzwords: dedicated data product managers (not side-of-desk), ontologists treated as first-class citizens, a VP-accountable governance process, and much more. See omnystudio.com/listener for privacy information.

Why multilingual content is a business risk, a compliance question, and an AI-readiness problem - not just a translation checkbox - with ServiceNow's Lyena Solomon. #ServiceNow #Localization #Globalization #AI Chapters 00:00  Cold open: "half ten" and the meaning problem00:29  Welcome + episode topic00:53  Meet Lyena Solomon01:12  Why this isn't just translation03:49  The word "order" - context matters05:05  What is language governance?06:04  Translating "pizza"07:03  The regulatory reality (Quebec, EU AI Act)09:30  Self-localization: Maori and Inuktitut13:23  The real business risk of inconsistency15:43  AI readiness and language risk16:28  A support ticket in three languages20:32  It's about trust, not just translation21:11  Closing thought: the joy of understanding22:17  Wrap-up + subscribeFor more about ServiceNow - https://www.youtube.com/@ServiceNowDocsTo watch these episodes on YouTube - https://www.youtube.com/watch?v=yxoHmZj5gOk&list=PLCOmiTb5WX3qvGq7Cp3o2KkCiplJyqQOK  See omnystudio.com/listener for privacy information.

This is the start of our 7th year of doing the podcast. Juan and Tim kick off the 12th season sharing who are the upcoming guests and what's the latest on their mind about this crazy data and AI world: The new website (honestnobsdata.com) and discussed building a knowledge graph of the podcast itself  Two types of semantics: for AI context vs. for system interoperability  The vendor hype gap vs. practitioner reality AI agent proliferation and the governance surface area explosion AI cybersecurity: models escaping their sandboxes  The economics of AI: accuracy, latency, and token cost as the new CAP theorem The great convergence, "triangle of consolidation" and “postmodern data stack” OSI and Apache Nessie Work as the center of gravity Bridging the operational and analytics worlds See omnystudio.com/listener for privacy information.

Voice AI sounds simple until you try to deploy it at scale — across airports, accents, languages, and thousands of employees at once. In this episode of ServiceNow Insights, host Bobby Brill sits down with Midam Kim, an ML engineer and linguist at ServiceNow, to unpack what it actually takes to build voice AI as an enterprise product. From the out-of-vocabulary problem (why AI still struggles with names) to why turn-taking in conversation is a linguistic skill most people never think about, Midam breaks down the human science behind the technology.  In this episode: - Why voice AI is replacing typing as the default way to interact with enterprise systems  - The difference between building for employees (B2B) vs. building for their customers (B2B2C)  - Why an airport is one of the hardest possible environments for voice AI — and what ServiceNow does about it  - The "out-of-vocabulary" problem: why AI still struggles with names, accents, and rare expressions  - Why ServiceNow's secret sauce is hiring linguists, not just engineers  - The linguistic framework behind every voice interaction: sounds, words, and turn-taking  - Why voice AI is like teaching a kid to speak for the first time Chapters00:00 — Welcome to ServiceNow Insights00:22 — Meet Midam Kim, ML Engineer & Linguist00:35 — Why voice is replacing typing01:48 — What voice AI actually does for employees03:40 — B2B vs. B2B2C: who's really using this?05:11 — Desk employee vs. airport traveler: two different problems06:42 — Building for an ever-changing environment08:53 — Why airports are the hardest use case09:54 — Accents, fluency, and the diversity problem11:20 — "My Name Is. My Name Is. My Name Is." — the OOV problem12:50 — The coffee shop name story13:20 — How ServiceNow trains its models14:59 — The 3 linguistic layers: sounds, words, interaction16:38 — Midam's turn-taking story from Korea18:33 — Why voice agents can't be "that person you avoid"20:16 — "We can make it great"  Subscribe for more ServiceNow Insights episodes on AI, voice technology, and enterprise innovation. Related episode: Voice AI Agent Evaluation — how ServiceNow measures whether voice AI meets human expectations. https://youtu.be/x7Ks932T18o For more about voice in AI from Midam Kim - https://youtu.be/3NUf6W_FMWs?is=wGc7BfyhiDp8JlOW  #VoiceAI #EnterpriseAI #ServiceNow #ArtificialIntelligence #Linguistics #ConversationalAI #AIProduct #PodcastSee omnystudio.com/listener for privacy information.

Most organizations deploying AI agents can’t answer a basic question: is it actually working? Not whether the agent runs — whether the process actually got better. In Episode 3 of our process mining and process intelligence series, Damian Pascale and Roz Parpia join host Bobby Brill to go deep on what it actually looks like to run process intelligence with AI in the mix — from the AI Visibility Gap, to a four-step framework for finding the right AI use cases, to what “closed loop intelligence” really means once agents are governing agents. This episode’s answer to the recurring question: don’t automate the chaos. Find it, understand it, improve it — then, and only then, streamline it.   CHAPTERS 0:00  Introduction — Damian Pascale & Roz Parpia 0:56  “Don’t automate the chaos” — where the phrase comes from 2:25  Process mining vs. process intelligence — what actually changed 4:21  The linchpin: where AI fits across all three layers 5:36  The AI Visibility Gap — what most organizations are missing 6:56  A real example: when agent metrics look great, but quality doesn’t 8:13  The four-step framework: Find, Understand, Improve, Streamline 11:27  Where AI comes into streamlining — sizing the right use cases 13:19  Does the order of the four steps actually matter? 14:04  Task Mining — the human side process mining can’t see 15:19  A concrete example: the procurement approval bottleneck 16:31  The closed loop — six steps to continuous improvement 17:30  Why you can never skip the ‘detect’ step 18:05  Measuring real impact with the compare feature 19:25  Governance and AI Control Tower, explained simply 20:46  Mining the agents themselves — a third layer of visibility 21:36  Closed loop intelligence — the three layers, confirmed 22:48  Day one: what to do after deploying your first agent 23:38  Closing thoughts from both guests 24:26  Wrap-up   IN THIS EPISODE • Why an AI agent doesn’t fix a broken process — it just runs the broken process faster • The real difference between process mining and process intelligence: three layers in one • The AI Visibility Gap: why almost every customer has deployed an agent, but few can prove it’s working • A real customer example — an agent that improved response time but quietly increased the reopen rate • The four-step framework for AI-ready process improvement: Find, Understand, Improve, Streamline • The 2–15 minute rule (and the 3–9 minute sweet spot) for sizing the right AI agent use cases • Why skipping straight to automation is exactly how you end up automating the chaos • Task Mining and the procurement approval example — 45 minutes across four systems, invisible to process mining alone • The six steps of the closed loop, and why the ‘detect’ step is the one everyone skips • Using the compare feature to measure whether an AI agent actually helped — or just moved the problem • AI Control Tower, explained simply — and how it becomes a third layer of process intelligence • Closed loop intelligence: the agent, the governance, and the agent’s own behavior — all observable, all improving   GET STARTED If you’re a ServiceNow customer, you already have access to free evaluation projects — no license needed. https://www.servicenow.com/au/products/process-mining/get-started.html https://www.servicenow.com/docs/r/now-intelligence/process-mining/process-mining.html https://www.youtube.com/watch?v=TVrU0TQ7ldM https://www.youtube.com/watch?v=GLKROYqnc10   #ServiceNow #ProcessMining #ProcessIntelligence #AIAgents #AgenticAI #TaskMining #AIGovernance #AIControlTower #WorkflowAutomation #DigitalTransformation #ContinuousImprovement #ClosedLoopIntelligence #ServiceNowPodcast #EnterpriseAI #DontAutomateTheChaosSee omnystudio.com/listener for privacy information.