The GOD Framework: How the Kirloskar Group filters tech hype from real value

Every technology leader has heard some version of the same pitch: this changes everything. Blockchain was going to rewrite supply chains. Then generative AI arrived promising to rewrite everything else. For most boardrooms, the challenge isn’t spotting these waves — it’s knowing which ones are worth riding and which will crash quietly a year later, leaving behind nothing but a stalled pilot and a line item nobody wants to explain.

At Kirloskar Group — a 138-year-old conglomerate spanning pumps, engines, chillers, compressors, electrical equipment, submarine pipelines, real estate, education, and non-banking financial services — that challenge is magnified by scale and diversity. There is no single “Kirloskar business” to optimize for; there are many, each with its own regulatory pressures, customer expectations, and definitions of value. Deciding what technology deserves investment across a portfolio this varied could easily become guesswork, driven by whichever business head shouts loudest in the boardroom.

Mandar Sahasrabudhe, Group CTO of Kirloskar Management Services Private Limited, has built a simpler filter instead — one he distills into a memorable acronym: GOD. Technology, in his framing, only ever does one of four things: Generate value, Operate as expected, Disrupt the status quo, or Disappear entirely once the hype fades. It’s a lens that turns an abstract, often emotional decision — should we chase this? — into a testable question with a concrete answer.

In practice, that means very little at Kirloskar gets adopted on conviction alone. New technologies, however fashionable, are first pushed into sandboxed proof-of-concepts designed to expose their real utility rather than their marketing promise. Sometimes the group goes further, deliberately sitting out a trend altogether — watching from a distance until the market separates genuine impact from noise before committing resources.

In this conversation, Mandar unpacks how that filtering discipline plays out in practice: how Kirloskar unifies IT strategy across fundamentally different businesses without forcing false uniformity, what “AI readiness” actually costs when your data lives across decades-old ERP systems and dispersed manufacturing plants, where digital twins and AI are already earning their keep on the shop floor, and how an Autodesk award-nominated chiller design reflects the same patient, outcome-first thinking at the engineering level. Underlying all of it is the same question the GOD framework was built to answer: not is this new, but does this actually work for us.

One Strategy, Many Businesses
Unifying IT across a group as sprawling as Kirloskar’s is, in Mandar’s words, “a balancing act between the technologies available vs the business need, budgets, compliance etc.” Rather than forcing every business unit onto identical applications, Kirloskar has drawn its unification line around shared infrastructure layers: a common data lake, BI and AI dashboarding, core infrastructure, and standardized tools for collaboration, productivity, and communication, along with ERP.

Where the group draws back from full standardization is at the application layer. With businesses spanning manufacturing, real estate, education, and NBFC (non-banking financial company) operations, Mandar explained that “the applications might not be unified as we span across multiple domains… thus giving rise to varied compliance, business requirements or priorities.”

That same rigor extends to Kirloskar Chillers Pvt. Ltd., a 20-year veteran of the HVAC&R industry, whose Screw Chiller Product and Series Design project was recognized as a finalist at Autodesk’s Imagine 2025 Awards.

The project’s origins were rooted in market opportunity. Kirloskar’s research identified strong demand for screw chillers in the pharmaceutical and chemical sectors — a market Mandar pegs at roughly ₹55 crore annually and growing at a compound annual growth rate of approximately 35%, with further industry expansion on the horizon. To capture that opportunity, the team needed to design a new chiller series capable of spanning capacities from 80TR to 540TR, using multiple combinations of heat exchangers (evaporators and condensers) and compressors.

The technical bar was high: “accuracy of design (3D model & 2D manufacturing drawing), interchangeability of component, product reliability and time to market” were, in Sahasrabudhe’s telling, the deciding factors for product success.

To hit that bar, the engineering team leaned heavily on Autodesk’s design ecosystem. Kirloskar Chillers uses Autodesk Inventor as its core 3D design platform and has built out design automation by linking Excel parameters with iLogic. For structural validation, the team uses Inventor Nastran to run linear static, non-linear static, modal, and frequency response analysis. The result, Mandar said, was a fully modular series with interchangeable components, where 3D CAD models and manufacturing drawings were generated through iLogic and reliability was verified across multiple finite element analysis (FEA) methods.

Data Readiness: The Unglamorous Phase Before AI
Mandar has been outspoken about a principle many enterprises learn the hard way: AI is only as good as the data foundation beneath it. For a conglomerate carrying legacy manufacturing equipment, decades-old ERP systems, and geographically dispersed plants, that foundation didn’t come cheap or fast.

“It actually was a complex journey of phase-wise implementation,” he said. The starting point was standardization — modernizing systems and establishing the right hybrid cloud strategy — followed by a common data lake wherever feasible, designed to consolidate dispersed data from IoT, OT, and IT systems through proper data modeling.

The hardest and longest part of the journey wasn’t the algorithms — it was the groundwork. “Data hygiene and management took most of the time,” Mandar said, “and then it was relatively simpler phase of BI/AI journey.”

Where AI Actually Pays Off in Heavy Engineering
Ask most technology leaders where AI delivers the highest ROI in manufacturing, and you’ll get a tidy answer: predictive maintenance, quality control, or supply chain optimization. Mandar’s answer is more nuanced — it depends entirely on the use case and the business objective behind it.

He distinguishes between direct and indirect returns: an AI deployment aimed at growing the spare parts business, for instance, produces a directly measurable ROI, while an efficiency-focused deployment tends to generate savings that are harder to attribute line by line.

The biggest payoff, in his view, comes not from any single function but from integration — AI use cases “integrated across these functions for seamless manufacturing operations,” driving efficiency, revenue growth, and faster product development and launch cycles.

Digital Twins: From Forecasting to Supply Chain
Digital twin technology has moved from buzzword to operational tool at Kirloskar. The group has implemented digital twins across its production lines, and Sahasrabudhe says the value has been concrete: stronger forecasting and predictive capability for production efficiency, with supply chain optimization the next frontier as the initiative matures.

Practically, that has translated into better decision-making on the shop floor — helping teams “make right decisions for corrective actions needed to meet production targets with desired quality and in desired time and cost.”

Tripling Revenue: Technology as Enabler, Not Sole Driver
Kirloskar has set an ambitious target: tripling revenue within four years through smart factory initiatives and smart processes. When asked how much of that growth is technology-enabled versus market-driven, Mandar resisted a tidy percentage split, arguing the two forces are linked indirectly rather than cleanly separable.

Technology, he explained, can accelerate product development or improve efficiency to meet market demand — but market demand can just as easily drive product innovation independent of any specific technology. What he was unambiguous about is the connective role technology plays across the whole equation: it is, in his words, “powering to deliver faster, ensuring right quality to meet the market demands.”

The Takeaway
For a conglomerate built over more than a century, Kirloskar’s technology playbook under Mandar reads less like a transformation story and more like a discipline: prove value before scaling, build data foundations before chasing AI, and let integration — not isolated pilots — be the real source of return. In an era when boardrooms everywhere are racing toward the next shiny platform, that patience may be the group’s quiet competitive edge.