DynPro's CEO on Solving Corporate Knowledge Loss with AI‑Driven Data Modernization

The market may be fixated on fully autonomous AI built from the group up, but that’s often not where the fastest or most sustainable ROI is emerging. The real value comes from something far more fundamental: getting legacy data into a state where AI can actually do its job. Enterprises focused on durable returns have realized a simple truth that autonomous agents can’t deliver lasting value without a modern, governed data foundation beneath them. The organizations investing in that foundation now are positioning themselves to win long term, because AI’s biggest bottleneck isn’t capability, it’s access to clean, interpretable, high-quality data. The leaders who remove that bottleneck early are quietly building an advantage that compounds with every new AI capability that follows.

We spoke with Shivkumar Thiagarajan, the Co-Chief Executive Officer at DynPro Inc., about the challenge of accessing, cleaning, and transforming data to drive ROI and efficiency gains. For Thiagarajan, this challenge is familiar territory. With 25 years of experience in management and business development, he has focused his team on using AI to solve this core data problem first. Their answer is an internal accelerator they call the "Data Modernization Genie," which zeroes in on data migration and illustrates AI's power as a force multiplier.

"The real challenge with AI and data is often a human one: the people who originally understood the old systems are no longer there, leaving business analysts unable to make sense of what’s in the data," Thiagarajan said.

• Ancient data, new speed: The difference isn’t just speed with AI, it’s also predictability and scale. The Genie automates much of the heavy lifting, reduces human error, and surfaces insights along the way, turning what was once a painstaking, high-risk process into a repeatable, efficient workflow. For enterprises burdened with decades of legacy systems, this acceleration doesn’t just save time, it enhances the ability to leverage AI on data that was previously inaccessible or incomprehensible. "Traditionally, data migrations took six months, but now we do them in weeks. With our Data Modernization Genie, we moved a 99-year-old company’s data into Snowflake. A project that initially took three months was most recently completed in just three weeks," Thiagarajan said.

Thiagarajan explained that the difficulty often stems from a deeply human problem: the loss of institutional expertise over time. As companies age, institutional knowledge evaporates as the individuals who built or were responsible for managing various infrastructure, systems, data, and workflow retire or leave the enterprise. With the original workforce who understood the old systems gone, it's a challenge for enterprises to surface any knowledge or SOP documentation, so their decisions live only inside tangled data structures. The problem is compounded when you consider the sheer volume of unstructured data legacy organizations possess.

• Corporate amnesia: Here, AI’s role shifts from an accelerator to a tool for digital archeology, sifting through disorganized data to rediscover business logic that has been forgotten. "The real challenge with AI and data is often a human one: the people who originally understood the old systems are no longer there, leaving business analysts unable to make sense of what’s in the data," Thiagarajan said.

• Three intelligent questions: "Analyzing large datasets is one area where AI shines especially bright. As Thiagarajan describes, AI can ingest massive data sets, analyze them, and generate three pointed and intelligent questions instead of a hundred. "Once a business analyst answers those, the AI builds a clean data dictionary automatically, leveraging both structured and unstructured data throughout customer environments," he said.

Yet even as AI continues to reveal new knowledge or increase workflow velocity, enterprise AI adoption lags for a simple reason: trust. The idea of unleashing an autonomous agent inside a core transactional system is a risk few leaders are willing to take. That hesitation is defining the pace of enterprise AI adoption, forcing a more evolutionary approach.

• A healthy fear: “AI adoption is an evolution because trust in AI agents is still a major hurdle. Most organizations are hesitant to give these agents full access to critical back-office systems like SAP or Salesforce,” Thiagarajan said. The solution, he argues, is not to force autonomy but to create safe operating zones within modern data platforms where AI has guardrails.

• A safe harbor for AI: "There’s a decisive role for modern data architecture, like Snowflake and Databricks, where data can be curated, controlled, and governed," Thiagarajan said. In these environments, organizations can structure and standardize data before AI ever touches it, providing a level of oversight and accountability that transactional systems alone cannot offer. “Once the data is in that governed state, AI agents perform remarkably well,” he said. This safe harbor strategy ensures that AI adds value without compromising control, giving leaders confidence to adopt transformative tools while maintaining oversight.

This is where partners and systems integrators step in. Their job is no longer just implementation but interpretation to begin bridging the gap between fast-moving AI innovation and the cautious reality of enterprise governance, helping CIOs make responsible decisions. Thiagarajan synthesizes this philosophy into a business model he called "AI-led services," which centers on providing a service augmented by AI-powered tools.

The path to AI adoption is a balancing act. Organizations want speed, but they also want control. They want outcomes, but they also want oversight. According to Thiagarajan, the most durable model is one where AI elevates human expertise rather than attempts to bypass it entirely. In his words, "We use AI agents to augment our team’s productivity, not to replace our people."

The path to AI adoption is ultimately a question of balance—between speed and control, automation and human judgment. The most successful enterprises deploy solutions that amplify their teams’ capabilities, making them smarter and more efficient while keeping human oversight firmly in place. Psychological safety is key: when leaders know that AI acts as an enabler rather than a replacement, they’re far more willing to embrace transformative technology. "The hesitation disappears when clients see that our accelerators are human-in-the-loop," Thiagarajan said. "AI helps teams move faster, but it doesn’t override their expertise. That balance is what turns curiosity into confidence, and experimentation into real, sustainable impact."