Digital transformation and enterprise system architecture expert delivers over 2.5 million dollars in savings and 150 percent revenue growth by stabilizing fragmented global systems despite the resistance of legacy operational models
As 2026 begins, the shift Gartner flagged last year is now the core enterprise AI direction: moving from copilots to task-specific agents built to execute workflows inside business systems. Gartner has also warned that a significant portion of agentic AI projects may be canceled by 2027 because of cost, unclear business value, and weak risk controls. That risk rises fast when agents are introduced into systems with weak governance, inconsistent data, and exception-heavy processes. If agents amplify the execution model you already have, what happens when that model is fragmented, under-governed, and full of hidden exceptions?
Agentic AI makes one requirement non-negotiable: Enterprise Resource Planning (ERP) has to be execution-ready before you delegate work to agents. Kasyapp Ivaaturi’s most relevant result to this theme is his work at Framestore, where he recovered an ERP implementation by aligning multiple business units and establishing a Center of Excellence. In an environment where AI is expected to trigger actions inside core systems, this kind of work is what makes outcomes repeatable across teams, keeps controls consistent, and reduces reliance on local exceptions. His coaching credentials reinforce the same operating discipline from the people side. He also brings formal change-leadership training that supports this kind of work. Ivaaturi is a Certified Intelligent Leadership Executive Coach (John Mattone). He also completed ICF-accredited leadership and executive coaching training through CCA Canada and is pursuing Professional Executive Coaching studies at the University of Texas at Dallas. These credentials help when ERP programs require clear decision ownership, consistent adoption across teams, and governance that holds while processes are being redesigned and automation expands.
Agentic AI in ERP signals a move from support to execution. Instead of suggesting next steps or summarizing data, it initiates actions inside workflows. It completes routine tasks end-to-end and escalates only the cases that require human judgment.
But the gap between promise and reality is growing. As Kasyapp says, “I often see teams build impressive agent demos, but they struggle when they roll them out across the company. The blocker is the ERP foundation. Decision ownership is unclear. Master data is inconsistent across units. Integrations fail under real load. Finance processes still depend on manual workarounds. When you put agents into that environment, they don’t make the process smoother. They repeat the same mistakes faster and at a larger scale.”
Ivaaturi’s current role as Vice President of Enterprise Applications at Framestore is a live example of what “execution-ready” looks like when the stakes are real. Framestore operates as a global visual effects and animation studio, delivering work for film, episodic content, advertising, and immersive experiences, with complex production and financial workflows behind the scenes. He was brought in to revive the company’s transformation strategy and lead ERP vendor due diligence and selection. More importantly, he rebuilt the conditions required for success after an ERP failure.
Kasyapp conducted a thorough analysis of the implementation by engaging stakeholders across business units, then regained management and private equity partner trust by redesigning the solution through workshops and by building a Center of Excellence. This is the type of work that determines whether agentic automation becomes a production capability or remains a pilot. A CoE is not bureaucracy. It is a governance spine. It standardizes how decisions are made, how processes are owned, and how data is treated as a shared asset rather than a local preference.
As he shares, “It was exhausting. Every unit brought its own rules, its own approvals, and its own exceptions. After the earlier failure, people didn’t trust the system, so I kept walking into spreadsheets, side trackers, and ‘temporary’ workarounds that had become permanent. The hard part was getting teams to drop their local fixes and commit to one standard.”
In spite of the challenges, Kasyapp delivered a clear financial outcome during the reset. He saved over £200 000 by developing and implementing solutions in-house instead of relying on an external partner. In parallel, he identified process improvement opportunities for automation to enhance operational effectiveness. The result is not only a better ERP program. It is an execution model that can survive scale, audit pressure, and future automation.
Agentic automation turns finance into an execution surface. When agents start posting entries, routing approvals, and triggering payments, long-standing “small” issues become hard blockers: inconsistent master data, regional workarounds, and spreadsheet-driven FX logic. At a global scale, there is no room for interpretation. Numbers must reconcile the same way everywhere.
That is why multi-country finance programs provide a clear test of whether an ERP core is ready for automation. At a global agricultural research organization, Kasyapp Ivaaturi led an upgrade from Microsoft Dynamics AX 2012 (an older enterprise resource planning system) to Microsoft Dynamics 365 Finance and Operations (a modern cloud platform used to run finance and supply chain workflows). The program supported a $220M annual institutional budget across four legal entities, 13 regional offices, four research stations, and projects in 40+ countries.
"In a global setup, everyone has their 'special' way of doing things, usually hidden in a localized spreadsheet. I kept telling the teams: we aren't just installing software; we’re building a single source of truth. It was messy work: standardizing math across 13 countries while ensuring a researcher in a remote station could still function. But you can’t automate a process that hasn't been disciplined first," the expert shares.
He redesigned finance and supply chain processes, generating huge savings. He integrated daily exchange rates into the ERP using sources that included Oanda and central bank references, replacing spreadsheet assumptions with consistent system inputs. He also designed and implemented a module managing $50 million across 40 countries that not only saved funds but also built credibility with the donors, and he led a shift in collaborator expense accounting from cash to accrual, supported by workshops and training for more than 200 users.
This is the foundation agents require: consistent process rules, clean data, and integrations that remove manual logic. Without that, automation only scales errors.
Execution readiness is often discussed in the context of finance operations and ERP programs. The same requirement applies in revenue-facing work and in large-scale public systems, where teams measure reliability through real operational throughput and financial control.
In commercial roles, Kasyapp Ivaaturi delivered results with direct financial impact. For example, at Nexus, he exceeded targets by 150% through consultative selling. He also led a five-year strategic plan that reduced consulting expenses by $1.4 million through company-wide cost initiatives.
He also managed large-scale delivery where systems must perform consistently at high volume. At NTT Data (a top-10 global IT services provider headquartered in Japan, which had acquired the US-based firm Keane), he was part of a $367 million contract to deliver a ticketing system for the Australian state of Victoria. The system processed one billion trips and $4 billion in funds, and the program reduced ticket stock handling (the logistics and accounting of physical tickets) and related costs.
"When you sync millions of smart-card handshakes across a fleet of moving buses and trains with different cellular dead zones, the math has to be bulletproof. One tiny error in the fare-calculation engine messes up a report and causes a line of five hundred angry commuters at a station gate. We had to move from a world of manual ticket reconciliations to a system where the data integrity was high enough to settle big sums of money without a human in the loop," Kasyapp Ivaaturi notes.
Such examples show the same underlying requirement across different contexts. Teams need clear ownership, controlled processes, and systems that maintain performance and accuracy under scale.
If agents are expected to trigger actions inside business systems, then governance becomes the limiting factor. Many projects fail when automation expands faster than ownership, controls, and data standards.
Kasyapp Ivaaturi’s approach reflects a sequencing logic that treats execution readiness not as a byproduct, but as the strategy itself:
"The sequence of implementation is more important than the technology itself. You cannot automate a system that is not yet under control," Ivaaturi notes. "My approach is to first stabilize the operating model by clearly defining ownership, approval hierarchies, and accountability. Once that is in place, we standardize the data and processes to ensure that reporting is the same in London as they do in Melbourne."
By establishing this baseline of consistency, organizations can transition from manual oversight to system-driven reliability.
"After the data is clean, we build the integrations to eliminate manual workarounds. Only then do we introduce automation," he continues. "We provide agents that operate within strictly defined rules and handle exceptions explicitly. If you skip the stabilization phase, automation only serves to scale existing errors."
In large-scale transformations, leadership capability must be operational, not decorative. A foundation in executive governance and behavioral change backs Kasyapp’s technical execution. He is a Certified Intelligent Leadership Executive Coach, trained in John Mattone's methodology (widely recognized for coaching figures like Steve Jobs), which focuses on aligning an executive’s internal values with their organizational impact.
His expertise is further validated by International Coaching Federation (ICF) accredited training through the Certified Coaches Alliance (CCA) in Canada and the University of Texas at Dallas, where he is currently pursuing advanced professional executive coaching credentials. This academic and professional background, which includes an MBA in Finance & Entrepreneurship and the London Business School’s Entrepreneurial Edge program, allows him to bridge the gap between technical systems and human adoption.
In environments where accountability and decision rights determine the success of a project, this training provides the governance discipline that agentic automation requires.
"Technical systems are only as effective as the people who manage them," he explains. "You can build a perfect ERP core, but if you haven't coached the leadership team to embrace new accountability models and move away from manual control, the technology will fail during adoption. My goal is to prepare the human infrastructure to handle the speed of a digital shift."
If your ERP and finance stack still runs on invisible exceptions, agents will not eliminate the problem. They will scale it. The organizations that win with agentic AI will look surprisingly “traditional” at the core: disciplined governance, clean process ownership, and integrations that make analytics actionable.
In that sense, Kasyapp Ivaaturi’s work offers a clear lesson for 2026. The fastest path to agentic value is not chasing autonomy. It is earning it by making execution reliable first.