Results

From Data Chaos to Confident AI: How Overdose Built a Semantic Foundation on Google Cloud

From Data Chaos to Confident AI: How Overdose Built a Semantic Foundation on Google Cloud

Most enterprises don't have a data problem. They have a meaning problem.

That's the framing Mike Ni, analyst at Constellation Research, brought to an interview with Paul Pritchard, CEO of Overdose, at Google Cloud Next. Throughout their conversation, Pritchard unpacked exactly what it takes to get from scattered data sources to AI that actually moves the business forward.

The answer, it turns out, starts well before the AI.


The Real Problem: Everyone's Fighting Over the Truth

Overdose works with merchants globally as a high-growth digital commerce agency operating across dozens of client relationships simultaneously. The challenge they kept running into wasn't a lack of data. It was that every team had a different version of it.

"With those disparate data sources comes a whole bunch of confusion," Pritchard explained. "Every channel claiming the sale meant that there was a fight every day to figure out what the truth was."

CFOs were looking at one number. Marketing managers at another. Board members at a third. And the agency was spending enormous energy just reconciling the differences before anyone could get to the actual work of growing the business.

The turning point came when Overdose built a semantic layer powered by Looker and Google BigQuery. Not another dashboard. Not another BI tool. A single, grounded source of truth that every team, client, and agency alike could work from.

"The ability to now sit down with a client and understand what the imperative metrics are (profit, effective movement of stock, customer acquisition that turns into lifetime value) and give everyone the same information at the same time has been so impactful," Pritchard said.


From Month-Long Dashboards to Real-Time Action

Before Looker, Overdose's process looked like this: spend a month building a dashboard, present it at the end of the month, and spend the meeting looking backward at what had already happened.

"You can imagine what you're doing at that stage," Pritchard said. "Looking back on the performance of that month and comparing it year on year."

Now, that same insight is available in real time. Teams don't wait for a monthly review to act. They build action-oriented plans based on what's happening now. And when an insight surfaces, generative AI is there to do something with it immediately.

Overdose built a tool that takes the semantic layer's output and runs it through an autonomous action loop: insight → creative brief → creative assets → published to ad channels → performance tracked back through Looker → new assets created when ready.

"Our merchants can move faster to engage with their customers," Pritchard said. "That's the best thing that can come out of it."


AI That's Actually Grounded

This is where the conversation gets pointed. Pritchard is not shy about his view on AI deployed without a solid data foundation.

"If it's not grounded in data, you get confident hallucinations," he said. "And if you get confident hallucinations, your brand is at risk."

It's a line worth sitting with. Confident hallucinations (i.e. AI that sounds certain while being wrong) are arguably more dangerous than obvious errors, because they're harder to catch and easier to act on. For a commerce agency making real-time decisions about campaigns, creative, and spend, that's not a theoretical risk.

This is also, Pritchard argued, why so many AI proofs of concept never make it to production. The semantic foundation wasn't there from the start. The AI had no solid ground to stand on.

"Intent is everything," he said. "A POC gives you some sort of intention. But the reality is, unless it's in the real world, connecting up all of those parts of your business to allow you to act on the fly in real time, that is the magic."


The People Question

One of the most common concerns when AI enters a business is what it means for the people doing the work. Pritchard's framing is worth noting.

"Instead of thinking about the reduction that everyone talks about, we think about the empowerment it brings to our teams," he said. "Our teams are now talking to clients about impactful metrics, rather than thinking about operational toolsets."

The shift isn't fewer people, but better conversations. Teams that used to spend time wrangling data are now spending time on strategy. And as the AI handles more of the execution layer, the human value moves further up the stack.


The Advice for Leaders Starting This Journey

Ni asked Pritchard what data and commerce leaders need to know before embarking on this path. His answer was direct: Start with the truth.

"What we knew is that disparate data sources leading to inconsistent reporting. Every channel claiming the sale meant there was a fight every day. When you get to that truth, what's beyond it is absolutely exponentially beneficial. Start by trying to get to the truth. Once you've got that, anything's possible."

For Overdose, the partnership with Google Cloud and the GCP team was central to getting there. Looker's price point made it accessible to the mid-market merchants it serves. And the consistency it created across every client relationship became the foundation for something bigger: an agency whose institutional knowledge and context compound over time.

"The information we retain over time should be able to be used to continually inform, elevate, and create innovation for those businesses," Pritchard said. "Through our partnership with Google Cloud, we've been able to do that in a consistent way — generating more revenue, more margin, and more customer acquisition for our merchants."


Watch the Full Interview

Catch the full conversation between Mike Ni and Paul Pritchard from Google Cloud Next below.

👉 https://youtu.be/w-0yDAaqyXQ

Data to Decisions Future of Work Innovation & Product-led Growth Tech Optimization Matrix Commerce Agentic AI AI Data to Decisions Chief Data Officer Chief Executive Officer On

Built to Last in the Age of Agentic AI: Why Technology Is Ready—but Leadership Isn’t | DisrupTV Ep. 436

Built to Last in the Age of Agentic AI: Why Technology Is Ready—but Leadership Isn’t | DisrupTV Ep. 436

Built to Last in the Age of Agentic AI: Why Technology Is Ready—but Leadership Isn’t

On this episode of DisrupTV, hosts R “Ray” Wang and Vala Afshar sit down with Joe Kim, CEO of DRUID AI, and Paul McCarthy, author of The Fired Leader, to unpack a hard truth:

AI is ready. Most organizations are not.

The conversation explores two forces colliding in real time—agentic AI and disruptive leadership—and why the gap between them is now the biggest risk (and opportunity) for enterprises.

AI Has Moved From Experiment to Executive Mandate

AI is no longer a side project or innovation lab exercise. It’s now a CEO-level priority shaping strategy, operations, and competitive advantage.

As Joe Kim explains, the shift from proofs of concept to production changes everything:

  • AI decisions now impact revenue, cost structures, and governance
  • Agentic systems introduce autonomy into workflows
  • The stakes move from experimentation to accountability

Even more surprising: in some cases, government agencies are advancing faster than the private sector in adopting agentic AI.

The takeaway is clear—this is no longer about testing AI. It’s about running the business with it.

Agentic AI Changes the Rules Entirely

Traditional enterprise systems are deterministic: same input, same output.

Agentic AI is not.

These systems reason, decide, and act—often without direct human intervention. That means organizations must rethink how they design, manage, and secure technology.

Instead of tools, agents must be treated like operators:

  • They have defined responsibilities
  • They require oversight and training
  • They can be manipulated or exploited

This shift introduces new risks, especially around security and control. Social engineering won’t just target humans—it will target AI agents.

To manage this, organizations need a control plane above the models: a layer that governs decisions, monitors behavior, and enforces guardrails.

What “Built-to-Last” AI Actually Looks Like

According to Joe Kim, durable AI systems share a few critical characteristics:

Modular architecture
AI stacks must be flexible. Models, agents, and data layers should be interchangeable—not locked together.

Data-first design
Clean, governed, and well-routed data remains the foundation of performance.

Observability and control
Organizations must be able to track, evaluate, and adjust agent behavior in real time.

Lifecycle management
Agents need versioning, testing, and promotion pipelines—just like software.

Cost discipline
Token usage must translate into measurable business outcomes, not just experimentation.

Guardrails by design
Without strong controls, agentic systems can behave unpredictably or unsafely.

In short: AI infrastructure is no longer just about capability—it’s about control, economics, and trust.

AI Is Becoming a Commodity—Execution Wins

As Ray points out, AI itself is rapidly commoditizing.

The advantage no longer lies in access to models, but in:

  • Distribution
  • Customer relationships
  • Integration into workflows

Incumbents now have a real opportunity to reinvent themselves—if they move fast.

If they don’t, the same forces will disrupt them.

The Real Bottleneck: Human Decision-Making

AI accelerates everything—except leadership.

With answers now generated in seconds, the limiting factor becomes:

  • How quickly leaders can interpret and decide
  • How comfortable they are with uncertainty
  • How well they manage non-deterministic systems

This shift is already creating pressure at the top. Many leaders are unprepared for the speed and ambiguity of AI-driven decision environments.

Transformation Is Not Technical—It’s Relational

Vala highlights a critical point: AI transformation is fundamentally about people.

Organizations adopting hundreds of agents are also:

  • Reskilling employees
  • Redesigning workflows
  • Reallocating talent internally

This isn’t just automation—it’s a redefinition of how humans and machines work together.

And that’s where leadership becomes the real constraint.

The FIRE Framework: What Future Leaders Need

Paul’s FIRE framework defines the traits organizations must cultivate:

  • Fresh thinking – challenging outdated assumptions
  • Inquisitiveness – asking better, deeper questions
  • Real accountability – authenticity with ownership
  • Expressiveness – willingness to challenge the status quo

These are exactly the qualities needed to lead in an AI-driven world—and the ones most organizations suppress.

From Cooks to Chefs

Vala offers a simple but powerful analogy:

Most companies hire “cooks”—people who follow recipes.

What they need are “chefs”—people who understand principles and can create something new.

In an AI economy:

  • Cooks scale existing systems
  • Chefs reinvent them

The companies that win will be those that cultivate—and protect—chef-like leaders.

Key Takeaways

  • AI has moved from experimentation to a CEO-level mandate
  • Agentic AI introduces autonomy, requiring new governance and control models
  • Durable AI systems depend on modularity, observability, and cost discipline
  • AI is commoditizing—execution and distribution now matter most
  • The biggest bottleneck is human decision-making speed and adaptability
  • AI transformation is fundamentally a people and relationship challenge
  • Most leadership systems are not designed for disruption
  • Organizations must identify and support leaders who challenge the status quo

Final Thoughts

The technology is ready.

Agentic AI can already automate, augment, and scale decisions in ways that were unimaginable just a few years ago.

But leadership hasn’t caught up.

The organizations that will define the next decade aren’t the ones with the most advanced models—they’re the ones that:

  • Build flexible, governed AI systems
  • Move faster in decision-making
  • And most importantly, empower leaders who think differently

Because in the age of AI, the constraint is no longer what machines can do.

It’s what leaders are willing—and able—to do with them.

Related Episodes

If you found Episode 436 valuable, here are a few others that align in theme or extend similar conversations:

Tech Optimization New C-Suite Digital Safety, Privacy & Cybersecurity AI Chief Executive Officer Chief Technology Officer Chief AI Officer Chief Experience Officer

From “activation energy” and agent orchestration to donkeycorns and relationship capital, DisrupTV 425 explains what actually separates AI hype from real business impact.

On DisrupTV <iframe width="560" height="315" src="https://www.youtube.com/embed/J5CupyHoVng?si=GHB_W8FIfV0UhJBW" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>

CRM Isn't Dead, Autonomous IT Is Here & Your 2026 Infrastructure Problem

CRM Isn't Dead, Autonomous IT Is Here & Your 2026 Infrastructure Problem

The latest episode of ConstellationTV features co-host analysts Holger Mueller and Liz Miller, unpacking developments in enterprise AI, marketing automation, CRM evolution, and vertical-specific technology solutions. This episode provides a roadmap for navigating today’s fast-moving, competitive tech landscape.


#1 Enterprise News Spotlight: Marketing Automation and AI Take Center Stage

Constellation analyst Liz Miller predicts a tidal wave of announcements signaling an evolutionary shift in how small and medium businesses access and use customer data platforms (CDPs) and AI. Companies like HubSpot, Canva, and Adobe are driving this change, with notable announcements such as HubSpot’s AI tools and AI-driven automation enhancements integrated into Canva’s newly acquired CDPs.

Miller anticipates a new marketing landscape: one no longer revolving around customer relationship management systems (CRMs) as the centerpiece, but instead anchored by CDPs that empower businesses to use AI-enhanced tools for real-time data and actionable insights.

“We’re gonna suddenly see the marketing automation space change from a very CRM-centric point of view... shifting to AI-powered platforms that prioritize flexible data accessibility.”

This signal shift is underpinned by broader trends, including the growing necessity for personalization, automation, and data-driven decision-making at scale. It represents a significant departure from traditional CRM-centric solutions towards platforms designed to harness the power of generative AI in everyday business processes.


CRM Redefined: Is CRM Dead or Evolving?

The often-repeated claim that CRM is “dead” is akin to the enduring existence of instant coffee and paper. Nevertheless, the role of CRM is undoubtedly changing as the center of gravity shifts away from reliance on capturing transactional data alone.

Miller notes that AI is redefining CRM by extending its capabilities to manage more flexible data and generate real-time insights. While CRMs remain critical for transactional records, AI is stepping in to drive contextual understanding and actionable intelligence that goes beyond traditional use cases. She asserts that platforms are finally settling into their true roles: “The right sizing of CRM is actually starting to happen.”

This evolution marks a broader trend in enterprise tech: the relationship between AI and legacy platforms. Rather than undermining CRM, AI complements these established systems, creating new value through dynamic decision support and deeper customer engagement opportunities.


The Enterprise AI Model Wars: Responsible Deployment and Competitive Strategy

Holger Mueller shifts the focus to the intense competition between enterprise AI providers, particularly OpenAI and Anthropic. He identifies major developments in Anthropic’s release process for its powerful AI models as an approach resembling safety standards for autonomous vehicles. By prioritizing responsible innovation, Anthropic demonstrates how enterprise AI needs guardrails to align cutting-edge technology with real-world operational demands.

With the introduction of Cursor, a coding-specific model designed for enterprise use, Holger notes OpenAI is racing to catch up with Anthropic in developing capabilities suited to closed systems—a dynamic that is sparking innovation and competition across the enterprise AI vendor ecosystem.

This rivalry underscores a broader theme: the demand for trustworthy AI systems tailored to operational realities. Mature deployment strategies, paired with an industry-wide shift towards responsible innovation, are crucial for ensuring AI becomes an enabler rather than a disruptor in enterprise contexts


Enterprise Infrastructure Evolution: Oracle’s Moves Toward Agentic AI

Holger discusses Oracle's advancements in agentic AI, cloud systems, and backend operational capabilities. These include the company’s cutting-edge technology for sub-second availability, which is a feat driven by its emphasis on massive scale and high reliability.

This is a proactive step by Oracle to prepare backend systems for the success of agentic AI. At its core, Oracle’s vision aligns with the idea that enterprise systems must evolve to accommodate emerging AI capabilities, providing robust infrastructure that supports automation, scalability, and real-time decision-making.

As enterprise vendors increasingly integrate AI into their offerings, Oracle’s forward-looking approach stands out for its focus on readiness, reliability, and cloud-first architecture.


Redefining AI Success Metrics and the OpenAI Conundrum

It is important to question how AI success is currently defined. Metrics must be rooted in practical use, rather than stock-market interpretations of tech hype. Highlighting OpenAI’s challenges, including the recent shutdown of its Sora platform, Liz warns against using unstable and overhyped platforms as industry benchmarks.

“We have to stop using the room that has no adult as the yardstick for every vendor’s success.”

Enterprises must remain skeptical of fleeting trendsetters and focus instead on sustainable, scalable, and value-driven AI innovation.


#2. Autonomous IT and Elastic’s Observability Innovations

Elastic has emerged as another significant player reshaping IT operations. In the next ConstellationTV segment, Constellation analyst Chirag Mehta describes Elastic's transition from traditional observability systems to newer autonomous IT platforms powered by AIOps and generative observability. The reason Elastic made Constellation's 2026 ShortList for top Autonomous IT Platforms.

These platforms go beyond displaying data. They actively produce insights and recommendations, enabling faster decision-making during incidents.

  • Speed and Control: Elastic’s agent model reflects the balance organizations need. “Teams want speed, but they also want control. Autonomy grows around guardrails and trust,” explains Mehta. Such frameworks will appeal to businesses seeking greater operational efficiency without sacrificing oversight. Elastic’s approach exemplifies how organizations can balance automation with human judgment.
  • Cost-Efficiency and Scalability: Elastic’s use of Elastic Search is a standout innovation. By efficiently organizing massive telemetry data, they mitigate the challenges posed by sprawling observability needs. As digital transformation continues to expand telemetry and AI usage, such cost-effective systems will be indispensable.

#3. Verticalized Platforms Lead the Way at Infor Analyst Summit

Finally, the episode shifts gears to examine the future of verticalized enterprise platforms, building on observations from the Infor Analyst Summit. A discussion with analysts R "Ray" Wang, Michael Ni, and Holger Mueller highlights the impact of vertical AI capabilities and industry-specific tools on enterprise growth.

Ni explains the importance of process transformation over data consolidation: “AI is getting a job, and that job is improving processes.” This evolution highlights the need for industry-specific solutions that go far beyond generic data tools to deliver transformative results. Wei adds that industry verticals, with their distinct needs and value chains, are increasingly driving AI strategies.

For example, to serve manufacturing, airlines, and other niches effectively, AI platforms—such as Infor’s Intelligent Applications—are focusing on integrating data lakes and transactional systems into unified conversational user experiences. As Muller notes, "The vertical depth matters more than breadth when it comes to digital transformation."

AI Applications and Speed to Value

Ni and Wang highlight the speed at which AI transformations are translating into real-world outcomes. Within three weeks of deployment, organizations can achieve game-changing efficiency. The strength lies in platforms that are built with built-in models and process mining capabilities.

For enterprise leaders navigating AI adoption, the lesson is twofold:

  1. Focus on platforms with pre-baked models that interconnect seamlessly.
  2. Prioritize rapid implementation to achieve speed-to-value metrics

Final Thoughts: Balancing Optimism with Responsibility

There is a vital balance in enterprise tech: optimism for transformative AI and automation on one side, tempered by the need for maturity, responsibility, and context in deployment strategies.

Whether through marketing automation, CRM evolution, enterprise AI model wars, or verticalized platforms, the key lies in leveraging robust technologies while exercising caution amid hype-driven narratives.

For enterprises looking to stay ahead, the takeaways are clear: adopt AI and automation thoughtfully, focus on reliable infrastructure, explore industry-specific solutions, and redefine success metrics to reflect operational impact rather than hype.

Stay tuned to Constellation Research for more insights on emerging trends in enterprise tech, as these dynamic conversations set the stage for continuous innovation.

Future of Work Tech Optimization Innovation & Product-led Growth Marketing Transformation Next-Generation Customer Experience AI CRM Agentic AI Big Data Chief Marketing Officer On ConstellationTV

AI is finally getting a job | Infor Analyst Summit Takeaways

AI is finally getting a job | Infor Analyst Summit Takeaways

Constellation Research analysts are back from the Infor Analyst Summit in Atlanta with one clear message: AI that doesn’t understand your business processes won’t move the needle.

Infor is doubling down on:

  • Industry value chains across 60+ sub-verticals
  • A business domain/context graph that connects data, processes, and semantics
  • Embedded process mining, RPA, and AI on a single platform
  • Speed to value, including discrete manufacturers optimizing SLAs within just 3 weeks of go-live

Constellation’s R “Ray” Wang, Holger Mueller, and Michael Ni also highlight a major leadership shift:
CEOs are now owning the AI agenda, demanding accountability, measurability, and risk mitigation across critical business processes.

This isn’t about “buying a tool” anymore.
It’s about operationalizing AI on top of well-understood, deeply modeled processes.

Watch the full conversation to see how Infor is approaching AI, vertical depth, and enterprise transformation.

Future of Work Data to Decisions Digital Safety, Privacy & Cybersecurity New C-Suite AI Agentic AI Robotics Chief Executive Officer Off

Leading Through the Polycrisis: Security, AI, and the Rise of the Polymath CEO | DisrupTV Ep. 435

Leading Through the Polycrisis: Security, AI, and the Rise of the Polymath CEO | DisrupTV Ep. 435

Leading Through the Polycrisis: Security, AI, and the Rise of the Polymath CEO

In a world of converging crises—AI, geopolitics, cyber risk, and climate—leadership is being redefined in real time.

On DisrupTV Episode 435, hosts R “Ray” Wang and Vala Afshar brought together Paul Abbate (former FBI Deputy Director), Dr. David Bray (CEO, LeadDoAdapt Ventures), and Caroline Stokes (CEO coach and author) to explore what it takes to lead in an era of constant disruption.

The message was clear: the future belongs to leaders who can navigate complexity, design for resilience, and stay grounded in human purpose.

1. The Polycrisis Is Real—and Most Organizations Aren’t Ready

Paul Abbate drew from decades in national security to highlight a growing gap: organizations are underestimating their exposure to modern threats.

The biggest risks aren’t just external—they’re internal and systemic:

  • Insider threats: Data theft, espionage, and human error remain under-addressed
  • AI-powered misinformation: Reputation and market risks are accelerating
  • Rising attack sophistication: AI is enabling faster, cheaper, more scalable attacks
  • Organizational complacency: Security is still treated as a project, not a discipline

His advice is direct: security must become always-on, intelligence-driven, and integrated across the business.

2. Resilience Is the Missing Boardroom Capability

Dr. David Bray pushed the conversation further: risk is inevitable—resilience is optional.

Most boards still lack a formal focus on resilience, despite increasing volatility. The shift leaders must make:

  • From preventing risk → absorbing and adapting to it
  • From siloed security → integrated physical + cyber defense
  • From reactive response → proactive preparedness

He also flagged emerging risks many organizations overlook:

  • Hardware-level compromise before devices even reach users
  • Synthetic employees using AI-generated identities
  • Corporate espionage tactics once reserved for nation-states

The takeaway: resilience is no longer operational—it’s strategic.

3. Rethinking Incentives in the Age of AI Risk

As AI accelerates vulnerability discovery, organizations face a paradox: more visibility into risk can look like worse performance.

Bray argues leaders must rethink incentives:

  • Reward early detection, not just prevention
  • Encourage transparency, not blame
  • Recognize that more “found issues” often means better systems—not worse teams

In short, security teams shouldn’t be punished for seeing more clearly.

4. AI Governance Is Now a CEO Mandate

As AI becomes core to business strategy, CEOs are effectively becoming Chief AI Officers.

Bray outlined a new leadership playbook:

  • Empower “responsible heretics” who challenge assumptions
  • Anchor decisions in data, not instinct alone
  • Create moral space to debate long-term consequences
  • Define decision thresholds before crises hit
  • Build in pivot paths for when strategies fail

This is governance for a probabilistic, fast-moving world—where certainty is rare and adaptability is everything.

5. The Rise of the Polymath Leader

Caroline Stokes reframed the leadership challenge: we are entering the era of the polymath CEO.

Future leaders must:

  • Synthesize across domains (AI, climate, geopolitics, society)
  • Translate complexity into action
  • Continuously learn and adapt
  • Lead both humans and AI agents

This shift is already underway. Many CEOs are stepping down—not because they failed, but because the role itself has fundamentally changed.

Boards, too, must evolve:

  • Update skill sets to include AI and systems thinking
  • Redefine expectations of leadership
  • Use AI as a “third voice” in decision-making

Leadership is no longer about optimization—it’s about reinvention.

6. Purpose, Pressure, and the Human Side of Leadership

Amid all the complexity, Stokes emphasized a critical truth: leadership is becoming more human, not less.

Leaders today face:

  • A global loneliness and burnout crisis
  • Constant exposure to instability and disruption
  • The challenge of moving faster than their organizations can absorb

Her guidance:

  • Invest in personal resilience (coaching, reflection, support systems)
  • Use AI as a thinking partner—not just a productivity tool
  • Design environments that foster connection, purpose, and trust

Because while AI accelerates execution, humans still determine alignment, meaning, and culture.

Key Takeaways

  • The polycrisis is here: AI, cyber, geopolitics, and climate risks are converging—and compounding.
  • Resilience is a strategic capability: Organizations must design for disruption, not avoid it.
  • Security must evolve: Insider threats, synthetic identities, and AI-driven attacks require new models.
  • Incentives need a reset: Reward visibility, learning, and proactive defense—not just outcomes.
  • The CEO role is changing: Leaders must become polymathic, AI-literate, and systems-oriented.
  • Purpose still matters: In a machine-scale world, human connection, trust, and meaning are differentiators.

Final Thoughts

DisrupTV Episode 435 makes one thing clear: this is not a normal operating environment.

We are entering an era where:

  • Crises are continuous, not episodic
  • AI reshapes both opportunity and risk
  • Leadership requires both technical fluency and human depth

The leaders who succeed won’t be the most specialized—they’ll be the most adaptive.

They will connect disciplines, build resilient systems, and guide organizations through uncertainty with clarity and purpose.

In the age of polycrisis, the ultimate advantage is not just intelligence—it’s integration.

Related Episodes

If you found Episode 435 valuable, here are a few others that align in theme or extend similar conversations:

Tech Optimization New C-Suite Digital Safety, Privacy & Cybersecurity AI Chief Executive Officer Chief Technology Officer Chief AI Officer Chief Experience Officer

From “activation energy” and agent orchestration to donkeycorns and relationship capital, DisrupTV 425 explains what actually separates AI hype from real business impact.

On DisrupTV <iframe width="560" height="315" src="https://www.youtube.com/embed/J5CupyHoVng?si=GHB_W8FIfV0UhJBW" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>

ShortList Spotlight: Elastic’s Data-First Approach to Autonomous IT

ShortList Spotlight: Elastic’s Data-First Approach to Autonomous IT

In modern incident response, the problem isn’t a lack of data, but rather turning that data into fast, confident operational decisions.

In this ShortList Spotlight, Constellation Research explores how autonomous IT platforms are reshaping incident management by combining observability, AIOps, and SRE practices.

CR analyst Chirag Mehta highlights Elastic’s data-first approach, showing how Elastic Search and agent-driven workflows help teams:

  • Cut through alert noise and connect signals across distributed systems
  • Move beyond dashboards to insights, root causes, and next best actions
  • Use agents as operational teammates to summarize incidents, correlate telemetry, and propose investigation steps
  • Scale generative observability, where the platform doesn’t just show data—it helps produce understanding
  • Keep the data layer efficient while making the intelligence layer operationally useful

If you’re facing growing telemetry volumes, rising retention needs, and increasing AI-driven analysis, learn why the next wave of observability is about autonomy with guardrails: speed + control

Digital Safety, Privacy & Cybersecurity AI Chief Information Security Officer On ShortList Spotlights

AI Becomes Operational: The Board Quarterly Update

AI Becomes Operational: The Board Quarterly Update

The rapid adoption of artificial intelligence (AI) in enterprises is transforming industries, and Ray Wang and Esteban Kolsky from Constellation Research are leading the conversation on how organizations can harness AI effectively. Kolsky's recently published The Board Quarterly Review provides valuable insights into current enterprise technology trends and actionable frameworks for AI implementation.

"Expertise is now a commodity, but experience is not." Leveraging real-world experience has become imperative to guiding enterprises in making sustainable, impactful technology decisions.

Here are the points from the kickoff event to help you better understand the discussion and take away practical applications for your organization.


AI Implementation in Enterprises

There is an ongoing shift in enterprise AI from "model fascination to execution oversight." Wang and Kolsky provided a clear timeline:

“2026: this is the year that enterprise AI moves from potential to execution.”

Currently, 97% of organizations use AI in some capacity, but many remain stuck in experimentation. The pitfall of running multiple simultaneous pilots without clear governance was addressed, with notable examples of companies taking bold steps to transition from pilots to scaled deployments:

  • Johnson & Johnson: Reduced 13,000 AI pilots to 3 fully deployed solutions.
  • Walmart: Scaled AI efforts to just six or seven impactful projects.

For organizations, the takeaway is this: redesign, followed by governance, then scaling is the recommended approach. Organizations must integrate AI into business processes by redesigning workflows to accommodate it, rather than simply adding it to existing systems.


Execution Challenges and Success Factors

Execution issues are surfacing faster than technical barriers, underscoring the importance of people, processes, and change management. Citing insights from Boston Consulting Group, Kolsky articulated:

"70% of the value AI creates comes from people, process, and change management rather than algorithms."

Leadership must prioritize these areas to capture value effectively. For example:

  • Governance Frameworks: Teams must deploy frameworks that support the AI strategy, covering cybersecurity, compliance, people and process adaptation, and clear metrics.
  • Leadership Ownership: With many CEOs stepping in to manage AI strategy directly, leadership's role is increasingly pivotal.

Enterprise Adoption Path for AI

Tracing the lifecycle of AI adoption:

2023: Began with bounded production: introductory AI pilots under tight constraints.

Today: Enterprises realize the need for guardrails and governance.

2026 and Beyond: AI becomes embedded in infrastructure, transforming workflows across departments.

"It’s not AI first. It’s AI everywhere," emphasized Wang.


Actionable Framework for AI Deployment

To help organizations move from experimentation to scaled implementation, Kolsky presented a practical framework:

Five Key Questions Enterprises Need to Address:

  1. Where does AI's value come from? Organizations must assess areas where AI delivers tangible outcomes.
  2. What is a control plane? Governing AI means defining boundaries and accountability, ensuring alignment with business goals.
  3. How do you manage costs (tokenomics)? AI costs, particularly cloud-based expenses, must be closely monitored.
  4. Who owns decisions? Leadership across departments (CIOs, CTOs, CHROs, etc.) must take ownership of their segments' AI strategies.
  5. What are the external dependencies? Understanding reliance on hyperscalers, APIs, and consultants is critical for sustainability.

Role of Leadership in AI Strategy

There is significant evolution in the role of the Chief AI Officer (CAIO). This role has shifted from a tech-centric position to one that requires strategic coordination and oversight. Wang emphasized:

“The Chief AI Officer ensures that AI efforts stay aligned, tracks deployments, and streamlines innovation.”

Interestingly, many CEOs today are taking on the CAIO role themselves, reflecting AI’s importance in shaping overall business strategies.


Business-Driven Technology Adoption

A key principle highlighted was aligning technology with business objectives:

“The business requirement comes first, and then the technology follows.”

This approach ensures that investments in AI are purposeful and deliver measurable value. Mistakes in earlier technology adoptions came from prioritizing trends over strategic business needs.


Final Thoughts

Organizations must move beyond the hype and focus on disciplined execution in AI implementation. Wang and Kolsky challenged enterprises to answer tough questions about governance, leadership, and business-process redesign rather than relying on pilots or reactive technologies.

As AI becomes an inseparable part of the enterprise, the conversation shifts from “why AI?” to “how do we govern and scale AI for maximum impact?” This review offers a roadmap for enterprises navigating this shift, providing strategies and frameworks grounded in real business needs.

If you would like to know more, please don't hesitate to connect with Esteban Kolsky to discuss board trends and Constellation's board-level subscription service. For more insights, reach out to [email protected] or explore their LinkedIn community.


Key Quotes to Remember:

  • "Expertise is now a commodity, but experience is not."
  • "AI is done. It’s part of the enterprise going forward."
  • "The business requirement comes first, and then the technology follows."
  • "It’s not AI first. It’s AI everywhere."
Board Strategy AI Chief Executive Officer On

AI Hype vs. Human Reality: Why SaaS Isn’t Dying and Healthcare Is Rising | DisrupTV Ep. 434

AI Hype vs. Human Reality: Why SaaS Isn’t Dying and Healthcare Is Rising | DisrupTV Ep. 434

Human-Centered AI, Nursing’s Future, and the “Lunatics” Driving Innovation

On DisrupTV Episode 434, hosts R “Ray” Wang and Vala Afshar brought together three distinct perspectives—from healthcare, AI strategy, and enterprise journalism—to explore what actually matters in an AI-driven world.

Featuring Ena Hull (President & CEO, Ultimate Health School), Dr. Michael Wu (Chief AI Strategist, PROS), and Ron Miller (Editor at FastForward), the conversation cut through hype to reveal a clear throughline: AI may scale systems, but humans determine outcomes.

Nursing at a Breaking Point—and a Path Forward

Ena Hull painted a stark picture: nursing is both one of the most stable, AI-resistant careers and one of the most strained.

Demand is surging due to aging populations and post-COVID attrition, yet supply is constrained by faculty shortages, regulatory hurdles, and burnout. The issue isn’t lack of interest—it’s a bottlenecked pipeline.

Her solution: rethink the model.

  • Modernize education: Move from semester-based systems to modular, skills-based, faster pathways.
  • Unlock faculty capacity: Transition experienced (and burned-out) nurses into teaching roles.
  • Leverage simulation: Use AI-driven clinical simulations to reduce reliance on scarce in-person training slots.
  • Design for longevity: Embed wellness, stress management, and leadership training early.

The takeaway is clear: without redesigning how nurses are trained and supported, the system will continue to leak talent faster than it can replace it.

The “SaaSpocalypse” Isn’t Happening—But Change Is

Dr. Michael Wu challenged the idea that LLMs and agents will replace SaaS wholesale.

His argument is grounded in reality: today’s AI systems are probabilistic, not deterministic. That creates hard limits.

LLMs struggle with:

  • Consistency and repeatability
  • Explainability and auditability
  • Real-time performance at scale

That makes them a poor fit for domains like pricing, finance, and compliance-heavy decisions.

But disruption is happening—just selectively.

AI excels in areas like marketing, content, and workflows where:

  • Precision isn’t mission-critical
  • Speed and scale matter more than perfection

Wu’s guidance for leaders:

  • Experiment aggressively—but intelligently
  • Focus on repetitive, high-value, data-rich tasks
  • Use AI to augment, not replace, core systems

The real opportunity isn’t ripping out SaaS—it’s pairing deterministic systems with AI where each performs best.

Why Breakthroughs Come from “Lunatics”

One of the most compelling ideas came from Wu’s historical lens on innovation.

Breakthrough innovation doesn’t come from incremental improvement—it comes from compounding ideas, often sparked by unlikely collaborations.

Historically, progress accelerated when:

  • Builders (“how” people) and thinkers (“why” people) came together
  • Ideas were exchanged freely across disciplines
  • “Crazy” ideas were tolerated long enough to mature

These groups—once meeting in Enlightenment-era salons and even full-moon gatherings—were dubbed “lunatics.”

The modern implication:

Organizations must intentionally create spaces where:

  • Cross-functional thinkers collide
  • Unconventional ideas are explored
  • Experimentation is safe, not career-limiting

Before every breakthrough is an idea that sounds unreasonable.

AI Hype vs. Enterprise Reality

Ron Miller grounded the discussion with a journalist’s perspective from the front lines of enterprise tech.

Despite rapid innovation, real-world results are uneven:

  • Only about one-third of AI projects reach production with measurable ROI
  • Costs remain high, with pricing models still evolving
  • Enterprises are still figuring out how to map AI usage to business value

At the same time, clear value is emerging in:

  • Developer productivity
  • Research and knowledge work
  • Meeting prep and decision support

Miller’s key insight: AI is most effective today as an assistive tool, not an autonomous replacement.

And importantly, human networks—conferences, conversations, lived experience—remain a critical source of insight that AI simply cannot replicate.

Key Takeaways

  • Human-centered design wins: AI scales systems, but human judgment, care, and creativity remain irreplaceable.
  • Not all AI fits everywhere: LLMs are powerful, but their limitations make them unsuitable for many core enterprise systems.
  • Education and workforce models must evolve: Nursing highlights how outdated structures—not lack of demand—create systemic shortages.
  • Innovation requires collision: Breakthroughs happen when diverse thinkers connect and “unreasonable” ideas are explored.
  • Execution matters more than hype: Most AI projects still struggle—leaders must focus on measurable outcomes, not experimentation alone.

Final Thoughts

Episode 434 reinforces a critical truth: the future isn’t about choosing between humans and AI—it’s about orchestrating both effectively.

The organizations that win won’t be the ones that automate the fastest, but the ones that:

  • Redesign systems around real human needs
  • Apply AI where it truly adds value
  • And create cultures where bold, even “lunatic,” ideas can thrive

In a world of accelerating technology, human-centered strategy isn’t a constraint—it’s the advantage.

Related Episodes

If you found Episode 434 valuable, here are a few others that align in theme or extend similar conversations:

Future of Work Tech Optimization New C-Suite AI Chief Executive Officer Chief Technology Officer Chief AI Officer Chief Experience Officer

From “activation energy” and agent orchestration to donkeycorns and relationship capital, DisrupTV 425 explains what actually separates AI hype from real business impact.

On DisrupTV <iframe width="560" height="315" src="https://www.youtube.com/embed/J5CupyHoVng?si=GHB_W8FIfV0UhJBW" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>

R "Ray" Wang at Indiaspora Global AI Summit

R "Ray" Wang at Indiaspora Global AI Summit

Don't miss R "Ray" Wang's keynote highlights at the Indiaspora Global AI Summit.

The first movers who adopt this AI revolution, whether agentic or generative, will drive massive change. It's a winner-takes-all market.

The problem with AI is that it's capital-intensive, right now, creating a closed system with few players. We have a centralization model rather than a decentralization model.

But that is all going to change.

We are in the Age of Exponential Scale, and infinite possibilities abound. This means you have to be faster, better, OR cheaper; then faster AND better; then faster AND better AND cheaper.

Want the full keynote? Reach out to [email protected] to get access today.

Future of Work Tech Optimization Agentic AI AI GenerativeAI Off

Slackbot vs Salesforce, Gemma 4’s Open‑Source Push & HubSpot’s AI Pricing Bet

Slackbot vs Salesforce, Gemma 4’s Open‑Source Push & HubSpot’s AI Pricing Bet

In Episode 127 of ConstellationTV, Constellation Research analyst Martin Schneider and Editor-in-Chief Larry Dignan joined forces to dissect three critical topics impacting enterprise technology today: Slack’s new AI-powered capabilities, the rise of open-source AI models like Gemma 4, and HubSpot’s outcome-based AI pricing strategy. R "Ray" Wang, founder of Constellation Research, rounds out the discussion with an insightful spotlight on Smartsheet, a leading contender in the work coordination platform market.


Slack’s AI Expansion: Game-Changer or Incremental Upgrade?

Slack recently announced the rollout of 30 new features for Slackbot, its AI assistant. These enhancements aim to streamline workflows and improve collaboration, but are they worth the investment?

Key Highlights:

  • Salesforce’s $27 Billion Bet: Larry Dignan kicked off the segment by quipping, "Slack is changing the world. It's the best thing since sliced bread...maybe we'll finally see the value from $27,000,000,000 spent on a new UI for Salesforce, really." For executives managing CRM integration costs, this underscores the scale of Salesforce's wager on Slack’s inclusion in its ecosystem.
  • Headless Software & Flexible Architecture: Martin Schneider elaborated on Slackbot’s architecture, affirming its flexibility: "It's a big move and big noise and, you know, a giant step for headless kind, right? The foundations are there, and that's the important thing." For CIOs, this statement highlights the importance of investing in architecture that scales with evolving AI models. Larry added: "What Slackbot is really doing from an architecture standpoint is leapfrogging the large language models (LLMs)."
  • Immediate Concerns: Larry voiced skepticism about Slack’s iterative AI updates, stating, "I'm a little skeptical because, you know, we've been playing around with Slack for five years now, but maybe they finally got it right. Maybe it is the ultimate teammate we're looking for." This cautionary view highlights the importance of evaluating vendor-driven innovation against tangible business outcomes, a priority for CFOs optimizing tech budgets.


Open-Source AI Models: Enterprise Adoption Taking Shape

The discussion turned to the momentum of open-source AI, with models like Gemma 4 rising to prominence. For executives tasked with balancing innovation and cost control, this segment offered critical perspectives.==

Key Highlights:

  • Global Competition in AI: Larry Dignan lamented the current state of U.S. AI, noting: "The U.S. is just getting clocked by Chinese open-source models." This acts as a wake-up call for CxOs to ramp up investments in enterprise AI or risk falling behind global competitors.
  • Business Customization: Larry emphasized the enterprise importance of open-source models, saying, "Open source models matter a hell of a lot to the enterprise because you're going to have to customize these things. It's gonna be about domain knowledge, not necessarily your big master foundation model."
  • Cost and Innovation: Martin Schneider added, "What we've seen over the last couple of years is just a very expensive global proof of concept. Enterprises are going to need to leverage open source models to lower costs and enable internal innovation." For CTOs, this validates a strategic shift toward open-source technologies as cost-effective, scalable solutions.


HubSpot’s Outcome-Based Pricing: Changing the SaaS Game

HubSpot is breaking new ground with its outcome-based pricing model for AI agents, charging businesses based on results rather than flat subscription fees. This experiment could reshape SaaS pricing strategies across industries.

Key Highlights:

  • Details of the Pricing Model: Larry broke down HubSpot’s approach: "HubSpot is pricing per outcome. They are charging you 50¢ per resolved conversation for customer service agents and $1 per lead recommended for outreach."
  • Potential Backlash and Timeliness: Larry acknowledged risks to adoption, remarking, "Every time we've seen this idea floated, customers kind of revolt. This time might just stick because it might actually be the right time."
  • Strategic Positioning: Martin explained why choosing outcome-based models makes sense for specific AI use cases: "The value that we get from SaaS software isn't the value that we got two, three, four, five years ago. It's got to change or else these guys are going to get so commoditized down to just death, right?"

For CFOs, HubSpot’s pricing approach represents both an opportunity and a risk. On the one hand, outcome-based models directly align vendor charges with business improvements. On the other hand, they demand rigorous ROI analysis, as poorly targeted implementation could lead to higher costs without performance gains.


ShortList Spotlight on Smartsheet: Work Coordination Platforms Redefining Collaboration

After the news segment, R "Ray" Wang brought Smartsheet into the spotlight, which has earned a place on Constellation Research’s shortlist for work coordination platforms. The enterprise collaboration market is poised for explosive growth, and Smartsheet is a leading option.

Market Insights:

  • Industry Growth: Ray highlighted that the market for work coordination platforms is projected to grow to $46.3 billion by 2031, expanding at an annual growth rate of 11.4%. For CEOs keeping tabs on digital transformation strategies, this is a clear indicator that investments in these platforms are essential for boosting organizational effectiveness.

Platform Strengths:

  • Efficiency & Accountability: Ray framed Smartsheet as a category leader, stating, "Smartsheet helps employees better organize, track, optimize their productivity and effectiveness."
  • Core Features: He detailed the platform’s capabilities, including centralized task tracking, automated workflows, shared real-time views, and interactive dashboards: "Smartsheet coordinates centralized task tracking, automated notifications, and updates project progress while providing dashboard insights."
  • Integration with Major Platforms: Smartsheet’s compatibility with Slack, Microsoft, and Google Workspace further reinforces its appeal to enterprises seeking seamless interoperability.

For CIOs and COOs prioritizing operational efficiency, the case for Smartsheet is becoming increasingly compelling. The platform’s ability to align teams, automate workflows, and enhance visibility across projects directly addresses enterprise challenges around collaboration and productivity.


Final Takeaways

This episode of CRTV highlighted transformative shifts across enterprise technology. Whether it was Slack’s ambitious AI overhaul or the rise of new pricing models through HubSpot, the discussions underscored the need for executive leaders to align tech strategies with business outcomes.

  • AI Adoption vs. ROI: Executives should prioritize AI investments that drive tangible benefits beyond hype. As Larry pointed out, "mid-market companies don't care about AI for the sake of AI. They care about growth and returns."
  • Open Source as a Cost-Control Strategy: CTOs should explore how open-source AI solutions can lower operational costs while enabling domain-specific innovation.
  • Outcome-Based SaaS Pricing: CFOs must reevaluate how SaaS pricing models impact profitability, ensuring experimentation aligns with bottom-line benefits.
  • The Rise of Work Coordination Platforms: Smartsheet’s capabilities are a strong example of platforms addressing inefficiencies in collaboration and project tracking. With the market growing in size and importance, enterprises must accelerate adoption.


Catch the Next CRTV Episode in Two Weeks!

🔔 Don't forget to follow and subscribe to ConstellationTV across the following platforms:

YouTube, Spotify, LinkedIn, and X for more insights into enterprise technology trends.

Data to Decisions Future of Work Tech Optimization AI Agentic AI On ConstellationTV