As generative AI continues to dominate conversations across industries, companies are grappling with how to harness its transformative power while ensuring tangible results. In a recent keynote, Jason Hardy, CTO for AI at Hitachi Vantara, provided a deeply insightful roadmap for aligning data strategies with overarching business goals. With decades of experience leading AI and data innovation at one of the world’s leading technology corporations, Hardy’s guidance goes beyond theory – he offers pragmatic lessons for navigating today’s AI revolution.
This article distills Hardy’s insights, exploring the immense potential of generative AI, its pitfalls, and actionable strategies for organizations to align their data initiatives with business objectives.
The Transformative Promise of AI and Data Integration
Generative AI has emerged as a powerful force shaping the global economy, with Hardy highlighting its potential to generate a staggering $25.5 trillion in increased GDP worldwide. From leveraging predictive analytics for industries like transportation to creating innovative tools such as Hitachi’s "Time Machine", the possibilities seem unlimited. Yet, as Hardy reminds us, this transformation is far from automatic. Nearly 90% of today’s generative AI production pilots fail to reach implementation, and approximately 30% of projects are abandoned altogether.
The core challenge lies in intention and execution: How can businesses ensure their AI initiatives drive measurable outcomes while avoiding costly missteps? Hardy’s session outlined a series of guiding principles to help organizations overcome these obstacles.
sbb-itb-8058745
Key Lessons for Aligning Data and Business Goals
1. Be Outcome-Driven: Start with the "Why"
In the race to adopt AI, many businesses fall into the trap of implementing technology for its own sake. Hardy emphasized the importance of being "outcome first, AI second." Before deploying a single GPU or launching a pilot, organizations must clearly define the business problem they aim to solve.
- Define Success: Identify specific, measurable outcomes tied to business goals. What KPIs will indicate success?
- Avoid Science Experiments: Without a clear "why", AI initiatives risk becoming aimless explorations with little ROI.
Quote: "A lot of customers say, ‘I want AI,’ but they forget why they want it."
2. Think Big, Start Small, Move Fast
While the transformative potential of AI is undeniable, Hardy advocates for a practical approach: test small, peripheral use cases first, build momentum, and scale up. This avoids disruptions to core business functions while fostering a learning environment.
- Focus on Low-Hanging Fruit: Early wins can build organizational confidence and enthusiasm.
- Embrace Speed Over Perfection: AI systems don’t need to be flawless at the outset. Rapid iteration allows teams to refine models over time.
This strategy mirrors Hardy’s analogy of building a muscle: Begin with manageable exercises before tackling the heavy lifting of core business transformation.
3. AI Is a Team Sport
One of the session’s standout points was the emphasis on collaboration. AI projects require cross-functional teams that go beyond the IT or data science departments.
- Build a "Pit Crew": Like a Formula 1 team, successful AI implementation relies on input from diverse stakeholders – legal, HR, operations, and business lines.
- Diversity Drives Success: Different perspectives ensure a more robust and adaptable AI strategy.
4. Address the Data Problem Head-On
Data is the lifeblood of AI, but Hardy acknowledged that many organizations struggle with messy, unstructured datasets. Rather than waiting for a perfect data environment, he encouraged businesses to take a pragmatic approach.
- Isolate Key Data: Focus on the datasets required to solve the immediate problem rather than overhauling the entire system.
- Accommodate Imperfections: Build AI systems flexible enough to handle "noisy" or incomplete data without compromising too much on accuracy.
Quote: "Beware of garbage in, garbage out, but don’t let imperfect data paralyze progress."
5. Invest in Robust Infrastructure
AI workloads are resource-intensive, requiring scalable, cost-efficient infrastructure. Hardy noted a growing trend of repatriating AI workloads from the cloud to hybrid environments to control costs.
- Build an AI-Friendly Tech Stack: Ensure your infrastructure includes scalable GPUs, storage, and MLOps pipelines.
- Plan for Growth: A solid tech foundation minimizes the need for retooling as projects mature.
6. Train and Empower Users
Adoption is often the most overlooked aspect of AI transformation. Hardy stressed the importance of training end-users to understand and trust AI tools.
- Combat Fear and Resistance: People worry that AI will replace their jobs. Effective training helps dispel fears and maximize adoption.
- Design for Users: Engage end-users during development to foster enthusiasm and ensure usability.
7. Build Ethical and Trustworthy AI
Hardy highlighted the ethical implications of AI, particularly in critical systems like energy and transportation. Transparency, fairness, and privacy must remain at the forefront of AI development.
- Ensure Robustness and Reliability: Trust is built through consistent, reliable results.
- Prioritize Security: Protect sensitive data with enterprise-grade compliance mechanisms.
Quote: "The more trust you build into your systems, the more users will embrace them, and the more robust they’ll become."
8. Leverage Partnerships Without Losing Ownership
Partnerships are essential for scaling AI efforts, but Hardy cautioned against outsourcing strategy entirely.
- Offload Low-Value Tasks: Delegate non-strategic elements to partners while keeping core decision-making in-house.
- Maintain Accountability: Retain ownership to ensure alignment with long-term business goals.
9. Iterate and Learn from Failures
Failure is an inevitable part of the AI journey. Hardy encouraged organizations to treat failures as opportunities to refine their approach.
- Embrace the 10% Success Rate: Even a small percentage of AI projects that succeed can deliver outsized returns.
- Use Failures as Blueprints: Document lessons learned to improve future initiatives.
Real-World Applications: Hitachi’s AI Innovations
Hardy’s insights were grounded in Hitachi’s extensive experience developing AI solutions for industries ranging from rail transportation to agriculture. Highlights included:
- Smart Bananas: Using IoT and AI to improve sustainable farming in Australia.
- Predictive Analytics for Penske: Optimizing fleet maintenance and reducing failures.
- Time Machine AI: An innovative tool preserving historical data versions for compliance and decision-making.
These case studies underscore how a clear focus on outcomes and industry-specific challenges can drive meaningful AI adoption.
Key Takeaways
- Start with the "Why": Clearly define business goals before investing in AI.
- Focus on Early Wins: Target peripheral use cases to build momentum.
- AI Requires Collaboration: Assemble cross-functional teams for diverse perspectives.
- Data Isn’t Perfect – Work with It: Use only the data you need and design AI systems to handle imperfections.
- Invest in Your Infrastructure: Build scalable, hybrid environments for long-term success.
- Empower End-Users: Train staff to trust and effectively use AI tools.
- Ethics Matter: Build transparent, secure systems to foster trust.
- Learn from Failures: Use setbacks to refine strategy and scale successes.
Conclusion
Generative AI and its associated technologies hold immense potential to reshape industries, but success requires a disciplined, strategic approach. By focusing on outcomes, fostering collaboration, and addressing the unique challenges of data and infrastructure, businesses can position themselves to thrive in this rapidly evolving landscape.
Jason Hardy’s keynote provides a valuable playbook for any leader navigating the complexities of AI transformation. While the road may be fraught with challenges, the rewards for those who align their data strategies with business goals are well worth the effort. As Hardy puts it, "The juice is worth the squeeze."
Source: "Aligning Data Strategies with Business Objectives: Breaking Down Silos for Success" – Big Data & Analytics, YouTube, Mar 6, 2026 – https://www.youtube.com/watch?v=KopCW-dveyE
