Show Notes
AWS re:Invent has increasingly become a signal for where enterprise technology is actually headed, not just what’s being promised. In this conversation from the show floor in Las Vegas, Pankaj Kumar, Global VP of Strategic Alliances and Head of the AWS Business Group at Kyndryl, shared a grounded perspective on what it really takes to move agentic AI from pilot projects into real production environments. Rather than focusing on abstract AI potential, the discussion centered on execution, data gravity, and the operational realities faced by Global 2000 enterprises. Inside Partnering is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Why So Many AI Pilots Stall One of the core challenges Pankaj highlighted is something many partner leaders are seeing first-hand: an explosion of pilots with very little follow-through. “There’s a lot of money being poured into agentic AI that start with the pilot, but a very small percentage of those pilot projects actually see ROI in terms of real world executable projects.” The issue isn’t lack of tooling or ambition. It’s that most pilots don’t connect deeply enough to the systems that actually run the business. Without access to mission-critical environments and operational data, pilots remain isolated experiments. The Role of Mission-Critical Systems Kyndryl’s position in the market is unusual. The company manages complex, long-running systems for a majority of Fortune 100 enterprises, including a significant portion of the world’s managed mainframe environments. “We manage more than two to three decades of mission-critical systems for our customers. I think we have access to data which very few in the industry have.” That access becomes the foundation for making agentic AI useful. Instead of training agents on abstract datasets, Kyndryl works directly with the operational knowledge embedded in real systems, workflows, and infrastructure. The Enhanced Agentic AI Framework At re:Invent, Kyndryl highlighted its enhanced agentic AI framework, which combines several layers: * Deep operational knowledge from managed services * Proprietary platforms like Kyndryl Bridge with billions of data points * Agent builders developed by Kyndryl * Native integration with AWS services “That repository of knowledge that we have is the critical fuel that you require to make an agent AI project successful.” This approach is not positioned as a one-size-fits-all product. Instead, the framework is designed to be customized by region, industry, and customer maturity. From Assessment to Execution Another theme that surfaced repeatedly was execution discipline. Kyndryl works across the entire transformation lifecycle, from assessment and migration to modernization and operations. “We are able to translate those pilot projects into real world executable projects.” That translation is reflected in business outcomes. According to Pankaj, a meaningful portion of recent Kyndryl signings already include an agentic AI component, suggesting that customers are moving beyond experimentation. AWS as a Strategic GTM Partner The conversation also explored the depth of the AWS partnership. Kyndryl has aligned its go-to-market motion with how AWS now sells by industry rather than horizontal services. “Our sales motion and our collaboration with AWS is in tune with how AWS is going to the market, what solutions they are taking to the market.” This includes co-sell engagement across financial services, insurance, healthcare, automotive, and the public sector, where Kyndryl was a finalist for multiple AWS Partner of the Year awards. Industry-Specific Execution Matters One insight particularly relevant for partner leaders is the emphasis on regional and industry focus. “Something that works in Japan could not be the same thing that we take to continental Europe or Australia.” Rather than forcing global uniformity, Kyndryl and AWS customize plays based on local regul
More from Inside Partnering
Samrah Khan on partner specialization, faster customer outcomes, and Google Cloud’s evolving ecosystem strategy The system integrator business is undergoing a fundamental transformation. For years, the traditional model was relatively straightforward. Technology vendors sold platforms, customers selected implementation partners, and SIs delivered projects that could stretch across multiple years. AI is changing that equation. Customers want more than technology implementation. They’re looking for partners who can help define their AI strategies, identify business opportunities, develop specialized solutions, and deliver measurable results in months rather than years. In my latest Inside Partnering conversation, I welcomed back Samrah Khan, Director of Google Cloud SI/GSI Ecosystem North America, to explore how these changes are reshaping the partner ecosystem. Samrah leads Google’s SI/GSI partner relationships at a particularly interesting moment. Google Cloud’s business experienced an amazing 82% year-over-year growth in Q2! But the more interesting story isn’t the growth itself. It’s how the expectations of customers, technology providers, and partners are changing. Here are six insights from our conversation. 1. Customers Want a Unified AI Strategy, Not Another Technology Implementation One of Samrah’s clearest observations is that enterprise customers are approaching AI very differently from previous technology transitions. They’re no longer simply shopping for infrastructure or selecting a systems integrator after choosing a platform. Instead, they expect technology providers and partners to work together from the beginning, bringing a comprehensive perspective on business transformation. “Customers are looking at both the partners as well as the hyperscalers or technology providers to come in and look at the whole enterprise from A to Z.” Thanks for reading Inside Partnering! Subscribe for free to receive new posts and support my work.
For partner leaders, one of the hardest questions isn’t identifying your best-performing partners. Most organizations already know who they are. The harder question is: Who should you invest in next? That question sits at the center of my conversation with Kenneth Fox, CTO and founder of ChannelScaler. Kenneth has spent his career in channels and partnerships, and his perspective combines the technology side of partner management with the practical realities of running partner programs. Inside Partnering is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
At Sage FUTURE in San Francisco, I sat down with Dan Miller, EVP of Financials and ERP at Sage, to talk about how Sage is thinking about AI, ecosystem partnerships, and the future of ERP platforms. One of the clearest themes from the conversation was this: finance leaders are not resisting AI. They simply want transparency, accountability, and confidence in the outputs. For partner leaders, ISVs, and ecosystem builders, that creates a very different opportunity than the generic “AI transformation” messaging dominating the market right now. Dan framed it around trust. “What finance and operations professionals really want is confidence in the tools that they’re deploying are doing exactly what they are expected to do.” That requirement shapes how Sage is building both its AI capabilities and its broader partner ecosystem. Inside Partnering is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
This is a free preview of a paid episode. To hear more, visit insidepartnering.substack.com
At Sage FUTURES in San Francisco, Gretchen O’Hara delivered a clear message to partners: the market is shifting fast - and the opportunity is massive. But despite the noise around AI, one thing hasn’t changed. Partners are still at the center of how businesses win. “The channel is one of the most important things because our partners are the ones that have the deepest industry expertise and are closest to customers that can really solve their unique, unique business challenges.” AI may accelerate execution, but it doesn’t replace the need for domain expertise, customer proximity, and trusted advisory. In fact, it amplifies it. Inside Partnering is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.