Join the event trusted by enterprise leaders for nearly two decades. VB Transform brings together the people building real enterprise AI strategy. Learn more
This article is part of VentureBeats special issue, The Real Cost of AI: Performance, Efficiency and ROI at Scale. Read more from this special issue.
Three years after ChatGPT launched the generative AI era, most enterprises remain trapped in pilot purgatory. Despite billions in AI investments, the majority of corporate AI initiatives never escape the proof-of-concept phase, let alone generate measurable returns.
But a select group of Fortune 500 companies has cracked the code. Walmart, JPMorgan Chase, Novartis, General Electric, McKinsey, Uber and others have systematically moved AI from experimental innovation theater to production-grade systems delivering substantial ROIin some cases, generating over $1 billion in annual business value.
Their success isnt accidental. Its the result of deliberate governance models, disciplined budgeting strategies and fundamental cultural shifts that transform how organizations approach AI deployment. This isnt about having the best algorithms or the most data scientists. Its about building the institutional machinery that turns AI experiments into scalable business assets.
We see this as a pretty big inflection point, very similar to the internet, Walmarts VP of emerging technology Desir�e Gosby said at this weeks VB Transform event. Its as profound in terms of how were actually going to operate, how we actually do work.
The statistics are sobering. Industry research shows that 85% of AI projects never make it to production, and of those that do, fewer than half generate meaningful business value. The problem isnt technicalits organizational. Companies treat AI as a science experiment rather than a business capability.
AI is already cutting some product-development cycles by about 40 percent, letting companies ship and decide faster than ever, said Amy Hsuan, chief customer and revenue officer at Mixpanel. But only for companies that have moved beyond pilots to systematic deployment.
The failure patterns are predictable: scattered initiatives across business units, unclear success metrics, insufficient data infrastructure andmost criticallythe absence of governance frameworks that can manage AI at enterprise scale.
Initial evaluation is also something too many organizations overlook, Sendbird head of product Shailesh Nalawadi emphasized at this weeks VB Transform. Before you even start building [agentic AI], you should have an eval infrastructure in place. No one deploys to production without running unit tests. And I think a very simplistic way of thinking about eval is that its the unit test for your AI agent system.
Simply put, you cant build agents like other software, Writer CEO and co-founder May Habib said at VB Transform. They are categorically different in how theyre built, operated and improved, and the traditional software development life cycle doesnt cut it with adaptive systems.
Agents dont reliably follow rules, Habib said. They are outcome-driven. They interpret. They adapt. And the behavior really only emerges in real-world environments.
The companies that have succeeded share a remarkably consistent playbook. Through interviews with executives and analysis of their AI operations, eight critical elements emerge that distinguish pilot-phase experimentation from production-ready AI systems:
Every successful AI transformation begins with unambiguous leadership commitment. This isnt ceremonial sponsorshipits active governance that ties every AI initiative to specific business outcomes.
At Walmart, CEO Doug McMillon established five clear objectives for AI projects: enhancing customer experience, improving operations, accelerating decision-making, optimizing supply chains and driving innovation. No AI project gets funded without mapping to these strategic pillars.
It always comes back to basics, Gosby advised. Take a step back and first understand what problems do you really need to solve for your customers, for our associates. Where is there friction? Where is there manual work that you can now start to think differently about?
We dont want to just throw spaghetti at the wall, explained Anshu Bhardwaj, Walmarts SVP of Global Tech. Every AI project must target a specific business problem with measurable impact.
JPMorgan Chases Jamie Dimon takes a similar approach, calling AI critical to our future success while backing that rhetoric with concrete resource allocation. The bank has over 300 AI use cases in production precisely because leadership established clear governance from day one.
Practical implementation: Create an AI steering committee with C-level representation. Establish 3-5 strategic objectives for AI initiatives. Require every AI project to demonstrate clear alignment with these objectives before funding approval.
The companies that scale AI successfully dont build point solutionsthey build platforms. This architectural decision becomes the foundation for everything else.
Walmarts Element platform exemplifies this approach. Rather than allowing teams to build isolated AI applications, Element provides a unified machine learning infrastructure with built-in governance, compliance, security and ethical safeguards. This allows teams to plug in new AI capabilities quickly while maintaining enterprise-grade controls.
The vision with Element always has been, how do we have a tool that allows data scientists and engineers to fast track the development of AI models? Parvez Musani, Walmarts SVP of stores and online pickup and delivery technology, told VentureBeat in a recent interview.
He emphasized that they built Element to be model agnostic. For the use case or the query type that we are after, Element allows us to pick the best LLM out there in the most cost-effective manner.
JPMorgan Chase invested $2+ billion in cloud infrastructure specifically to support AI workloads, migrating 38% of applications to cloud environments optimized for machine learning. This wasnt just about compute powerit was about creating an architecture that could handle AI at scale.
Practical implementation: Invest in a centralized ML platform before scaling individual use cases. Include governance, monitoring, and compliance capabilities from day one. Budget 2-3x your initial estimates for infrastructurescaling AI requires substantial computational resources.
The most successful companies resist the temptation to pursue flashy AI applications in favor of high-ROI use cases with clear business metrics.
Novartis CEO Vas Narasimhan was candid about early AI challenges: Theres a lot of talk and very little in terms of actual delivery of impact in pharma AI. To address this, Novartis focused on specific problems where AI could deliver immediate value: clinical trial operations, financial forecasting, and sales optimization.
The results were dramatic. AI monitoring of clinical trials improved on-time enrollment and reduced costly delays. AI-based financial forecasting outperformed human predictions for product sales and cash flow. AI does a great job predicting our free cash flow, Narasimhan said. It does better than our internal people because it doesnt have the biases.
Practical implementation: Maintain an AI portfolio with no more than 5-7 active use cases initially. Prioritize problems that already cost (or could generate) seven figures annually. Establish clear success metrics and kill criteria for each initiative.
Traditional IT project structures break down when deploying AI at scale. Successful companies create AI podscross-functional teams that combine domain expertise, data engineering, MLOps and risk management.
McKinseys development of Lilli, its proprietary AI research assistant, illustrates this approach. The project started with three people but quickly expanded to over 70 experts across legal, cybersecurity, risk management, HR and technology.
The technology was the easy part, said Phil Hudelson, the partner overseeing platform development. The biggest challenge was to move quickly while bringing the right people to the table so that we could make this work throughout the firm.
This cross-functional approach ensured Lilli met strict data privacy standards, maintained client confidentiality, and could scale to thousands of consultants across 70 countries.
Practical implementation: Form AI pods with 5-8 people representing business, technology, risk, and compliance functions. Give each pod dedicated budget and executive sponsorship. Establish shared platforms and tools to prevent reinventing solutions across pods.
Enterprise AI deployment requires sophisticated risk management that goes far beyond model accuracy. The companies that scale successfully build governance frameworks that manage model drift, bias detection, regulatory compliance and ethical considerations.
JPMorgan Chase established rigorous model validation processes given its regulated environment. The bank developed proprietary AI platforms (including IndexGPT and LLM Suite) rather than relying on public AI services that might pose data privacy risks.
Walmart implements continuous model monitoring, testing for drift by comparing current AI outputs to baseline performance. They run A/B tests on AI-driven features and gather human feedback to ensure AI utility and precision remain high.
At the end of the day, its a measure of, are we delivering the benefit? Are we delivering the value that we expect, and then working back from there to basically figure out the right metrics? Gosby explained.
Practical implementation: Establish an AI risk committee with representation from legal, compliance, and business units. Implement automated model monitoring for drift, bias, and performance degradation. Create human-in-the-loop review processes for high-stakes decisions.
Perhaps the most underestimated aspect of AI scaling is organizational change management. Every successful company invested heavily in workforce development and cultural transformation.
JPMorgan Chase increased employee training hours by 500% from 2019 to 2023, with much of that focused on AI and technology upskilling. The bank now provides prompt engineering training to all new hires.
Novartis enrolled over 30,000 employeesmore than one-third of its workforcein digital skills programs ranging from data science basics to AI ethics within six months of launching the initiative.
This year, everyone coming in here will have prompt engineering training to get them ready for the AI of the future, said Mary Callahan Erdoes, CEO of JPMorgans asset & wealth management division.
Practical implementation: Allocate 15-20% of AI budgets to training and change management. Create AI literacy programs for all employees, not just technical staff. Establish internal AI communities of practice to share learnings and best practices.
The companies that scale AI successfully treat it like any other business investmentwith rigorous measurement, clear KPIs and regular portfolio reviews.
Walmart uses internal ROI calculations and sets specific metric checkpoints for teams. If an AI project isnt hitting its targets, they course-correct or halt it. This disciplined approach has enabled Walmart to scale successful pilots into hundreds of production AI deployments.
Our customers are trying to solve a problem for themselves, said Gosby. Same thing for our associates. Did we actually solve that problem with these new tools? This focus on problem resolution can drive measurable outcomes.
JPMorgan Chase measures AI initiatives against specific business metrics. The banks AI-driven improvements contributed to an estimated $220 million in incremental revenue in one year, with the firm on track to deliver over $1 billion in business value from AI annually.
Practical implementation: Establish baseline KPIs for every AI initiative before deployment. Implement A/B testing frameworks to measure AI impact against control groups. Conduct quarterly portfolio reviews to reallocate resources from underperforming to high-impact initiatives.
The most successful companies dont try to scale everything at once. They follow an iterative approach: prove value in one area, extract learnings, and systematically expand to new use cases.
GEs journey with predictive maintenance illustrates this approach. The company started with specific equipment types (wind turbines, medical scanners) where AI could prevent costly failures. After proving ROIachieving zero unanticipated failures and no downtime on certain equipmentGE expanded the approach across its industrial portfolio.
This iterative scaling allowed GE to refine its AI governance, improve its data infrastructure and build organizational confidence in AI-driven decision making.
Practical implementation: Plan for 2-3 scaling waves over 18-24 months. Use early deployments to refine governance processes and technical infrastructure. Document learnings and best practices to accelerate subsequent deployments.
The financial reality of scaling AI is more complex than most organizations anticipate. The companies that succeed budget for the full cost of enterprise AI deployment, not just the technology components.
But one thing to remember is that AI spending is more nuanced than traditional software, Groq CEO Jonathan Ross noted onstage at VB Transform. One of the things thats unusual about AI is that you cant spend more to get better results, he said. You cant just have a software application, say, Im going to spend twice as much to host my software, and applications can get better.
JPMorgan Chases $2+ billion investment in cloud infrastructure represents roughly 13% of its $15 billion annual technology budget. Walmarts multi-year investment in its Element platform required similar scalethough exact figures arent disclosed, industry estimates suggest $500 million to $1 billion for a platform supporting enterprise-wide AI deployment.
These investments pay for themselves through operational efficiency and new revenue opportunities. Walmarts AI-driven catalog improvements contributed to 21% e-commerce sales growth. JPMorgans AI initiatives are estimated to generate $1-1.5 billion in annual value through efficiency gains and improved services.
The human capital requirements for enterprise AI are substantial. JPMorgan Chase employs over 1,000 people in data management, including 900+ data scientists and 600+ ML engineers. Novartis invested in digital skills training for over 30,000 employees.
But these investments generate measurable returns. JPMorgans AI tools save analysts 2-4 hours daily on routine work. McKinsey consultants using the firms Lilli AI platform report 20% time savings in research and preparation tasks.
Often overlooked in AI budgeting are the substantial costs of governance, risk management and compliance. These typically represent 20-30% of total AI program costs but are essential for enterprise deployment.
McKinseys Lilli platform required 70+ experts across legal, cybersecurity, risk management, and HR to ensure enterprise readiness. JPMorgans AI governance includes dedicated model validation teams and continuous monitoring systems.
The most successful AI deployments are fundamentally about organizational transformation, not just technology implementation. The companies that scale AI successfully undergo cultural shifts that embed data-driven decision making into their operational DNA.
If youre adding value to their lives, helping them remove friction, helping them save money and live better, which is part of our mission, then the trust comes, Walmarts Gosby noted. When AI improves work, saves time and helps workers excel, adoption and trust follow.
The most successful companies dont treat AI as a specialist capability confined to data science teams. They embed AI literacy throughout the organization.
Novartis adopted an unbossed management philosophy, cutting bureaucracy to empower teams to innovate with AI tools. The companys broad engagement30,000+ employees enrolled in digital skills programsensured AI wasnt just understood by a few experts but trusted by managers across the company.
Rather than viewing AI as a replacement for human expertise, successful companies frame it as augmentation. JPMorgans Dimon has repeatedly emphasized that AI will augment and empower employees, not make them redundant.
This narrative, backed by retraining commitments, reduces resistance and encourages experimentation. GE ingrained AI into its engineering teams by upskilling domain engineers in analytics tools and forming cross-functional teams where data scientists worked directly with turbine experts.
The difference between pilot-phase AI and production-grade AI systems lies largely in governance. The companies that successfully scale AI have developed sophisticated governance frameworks that manage risk while enabling innovation.
Walmarts Element platform exemplifies the centralized platform, distributed innovation model. The platform provides unified infrastructure, governance, and compliance capabilities while allowing individual teams to develop and deploy AI applications rapidly.
This approach gives business units the flexibility to innovate while maintaining enterprise-grade controls. Teams can experiment with new AI use cases without rebuilding security, compliance, and monitoring capabilities from scratch.
The change that were seeing today is very similar to what weve seen when we went from monoliths to distributed systems, said Gosby. Were looking to take our existing infrastructure, break it down, and then recompose it into the agents that we want to be able to build. This standardization-first approach supports flexibility, with services built years ago now able to power agentic experiences through proper abstraction layers.
JPMorgan Chase implements risk-adjusted governance where AI applications receive different levels of scrutiny based on their potential impact. Customer-facing AI systems undergo more rigorous validation than internal analytical tools.
This tiered approach prevents governance from becoming a bottleneck while ensuring appropriate oversight for high-risk applications. The bank can deploy low-risk AI applications quickly while maintaining strict controls where needed.
All successful AI deployments include continuous monitoring that goes beyond technical performance to include business impact, ethical considerations and regulatory compliance.
Novartis implements continuous monitoring of its AI systems, tracking not just model accuracy but business outcomes like trial enrollment rates and forecasting precision. This enables rapid course correction when AI systems underperform or market conditions change.
The companies that successfully scale AI have developed sophisticated budgeting approaches that account for the full lifecycle costs of enterprise AI deployment.
Rather than funding individual AI projects, successful companies invest in platforms that support multiple use cases. Walmarts Element platform required substantial upfront investment but enables rapid deployment of new AI applications with minimal incremental costs.
This platform-first approach typically requires 60-70% of initial AI budgets but reduces the cost of subsequent deployments by 50-80%. The platform becomes a force multiplier for AI innovation across the organization.
JPMorgan Chase manages AI investments like a portfolio, balancing high-certainty, incremental improvements with higher-risk, transformational initiatives. This approach ensures steady returns while maintaining innovation capacity.
The bank allocates roughly 70% of AI investments to proven use cases with clear ROI and 30% to experimental initiatives with higher potential but greater uncertainty. This balance provides predictable returns while enabling breakthrough innovations.
Successful companies budget for the complete AI lifecycle, including initial development, deployment, monitoring, maintenance, and eventual retirement. These full-lifecycle costs are typically 3-5x initial development costs.
McKinseys Lilli platform required not just development costs but substantial ongoing investments in content updates, user training, governance, and technical maintenance. Planning for these costs from the beginning prevents budget shortfalls that can derail AI initiatives.
The companies that scale AI successfully use sophisticated measurement frameworks that go beyond technical metrics to capture business impact.
Walmart measures AI initiatives against business outcomes: e-commerce sales growth (21% increase attributed partly to AI-driven catalog improvements), operational efficiency gains, and customer satisfaction improvements.
JPMorgan Chase tracks AI impact through financial metrics: $220 million in incremental revenue from AI-driven personalization, 90% productivity improvements in document processing, and cost savings from automated compliance processes.
Beyond lagging financial indicators, successful companies track leading indicators that predict AI success. These include user adoption rates, data quality improvements, model performance trends, and organizational capability development.
Novartis tracks digital skills development across its workforce, monitoring how AI literacy correlates with improved business outcomes. This helps the company identify areas where additional training or support is needed before problems impact business results.
Companies that scale AI successfully manage their AI initiatives as a portfolio, tracking not just individual project success but overall portfolio performance and resource allocation efficiency.
GE evaluates its AI portfolio across multiple dimensions: technical performance, business impact, risk management, and strategic alignment. This enables sophisticated resource allocation decisions that optimize overall portfolio returns.
For enterprises looking to move from AI experimentation to scaled production systems, the experiences of these Fortune 500 leaders provide a clear roadmap:
- Establish an executive AI steering committee
- Define 3-5 strategic AI objectives aligned with business strategy
- Begin platform infrastructure planning and budgeting
- Conduct an organizational AI readiness assessment
- Implement a centralized AI platform with governance capabilities
- Launch 2-3 high-ROI pilot initiatives
- Begin workforce AI literacy programs
- Establish risk management and compliance frameworks
- Scale successful pilots to broader deployment
- Launch the second wave of AI initiatives
- Implement continuous monitoring and optimization processes
- Expand AI training and change management programs
- Integrate AI capabilities into core business processes
- Launch the third wave focusing on transformational use cases
- Establish AI centers of excellence
- Plan for next-generation AI capabilities
The enterprises that have successfully scaled AI share a common understanding: AI transformation is not primarily about technologyits about building organizational capabilities that can systematically deploy AI at scale while managing risk and generating measurable business value.
As Dimon observed, AI is going to change every job, but success requires more than good intentions. It demands disciplined governance, strategic investment, cultural transformation, and sophisticated measurement frameworks.
The companies profiled here have moved beyond the hype to create durable AI capabilities that generate substantial returns. Their experiences provide a practical playbook for organizations ready to make the journey from pilot to profit.
The window for competitive advantage through AI is narrowing. Organizations that delay systematic AI deployment risk being left behind by competitors who have already mastered the transition from experimentation to execution. The path is clearthe question is whether organizations have the discipline and commitment to follow it.
Daily insights on business use cases with VB DailyIf you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI.
Thanks for subscribing. Check out more VB newsletters here.
An error occured.
