Manish Sood and the trouble with being early

The awkward part of being early is that the customer may understand the problem perfectly and still decline your answer. In 2011, Manish Sood was making the case for bringing enterprise data together in the cloud. One large financial institution wanted the information unified. Putting its client records into a public-cloud service, however, was beyond what it could imagine accepting. Sood carried on building Reltio.

Fifteen years later, on May 7, 2026, SAP completed its acquisition of the company. The distance between those two moments contains a familiar Silicon Valley progression: software, customers, investment, expansion. Sood’s route through it is more particular. He began as a mechanical engineer, learned the frustrations of enterprise information from inside the industry, and put his own savings behind a different way of handling it.

The story also contains a lawyer whose example mattered before any investor arrived. She was his wife.

A question closer to home

Before Sood launched Reltio, his wife had already opened a solo law practice, just two years after law school. He later described how her willingness to proceed stayed with him. When he considered leaving the security of an established employer, she asked what the worst outcome might be. He thought about losing their savings. He reasoned that even then, he could get another job.

There is no need to make that exchange more cinematic than it was. Its force lies in the ordinary arithmetic: a household, accumulated savings, a professional career that could be resumed. Entrepreneurship looks rather different when the first investment has to come out of your own account.

For Sood, her example helped turn an industry observation into a decision. He had seen a persistent problem at work. Now he would have to discover whether he could build a company around fixing it.

An engineer learns where the stress goes

Sood grew up in India and graduated in mechanical engineering in 1995. His degree came from Andhra University College of Engineering. Software entered his education through practical work such as finite element analysis, using mathematical models to investigate where structures might fail. The computer was useful because the calculation was difficult and the answer mattered.

Early work on power-plant automation brought him closer to the systems companies used to run their operations. He later moved into enterprise technology, including years at Siperian and Informatica working on master data management. This was an education in a different kind of structural stress: information held in separate applications, with no easy agreement about the people, products, and organizations it described.

Informatica acquired Siperian in 2010. Sood went on to start Reltio the following year. He was entering a market whose frustrations he had already spent years trying to address. His wager concerned the architecture: cloud infrastructure and newer database approaches could support a more flexible service than the systems customers had been using.

The cloud arrives. The doubt stays.

In Sood’s account of Reltio’s beginnings, several changes were converging. Cloud infrastructure was becoming available to software companies. Internet businesses were demonstrating ways to distribute work across many machines. Meanwhile, enterprises were accumulating more applications, each producing another pocket of information.

That combination gave him a technical opening and a commercial problem. Adding capacity across machines offered a route beyond repeatedly buying a larger computer. Delivering software as a service could also change the burden of running and maintaining it. Neither advantage automatically settled a prospective customer’s doubts about where important records would live.

Sood’s own operating decisions followed his argument. When Reltio needed email, he resisted buying an office server. A company selling a service in the cloud should use that model itself. It was a modest purchase decision with a rather large principle attached. Founders sometimes announce their philosophy from a stage; occasionally they express it by refusing a piece of office equipment.

The early company was small and financed by Sood. Institutional capital came later, with the first financing closing around the turn of 2015. The order mattered: there was a business to examine before there was a venture round to celebrate.

A business built between applications

The phrase master data management does little to advertise its own usefulness. Consider the smaller irritation it is meant to prevent. You buy something online, speak to support, and visit a store. Each encounter can create a record. If the systems cannot connect those records, you get to explain yourself again. The company may have plenty of information about you while remaining surprisingly bad at recognizing you.

Sood has used retail journeys to explain why core data matters. A customer, a product, a supplier, a location: these are the recurring subjects of a company’s activity. The applications surrounding them may change. The need to identify them consistently persists.

In a 2021 community conversation, asked about his passion, he answered: “Data, data, and data!” It is an admirably economical answer from someone whose business involves eliminating duplication.

Reltio’s work brings records together, resolves matches, and makes connected information available to other systems. The timing matters as much as the tidiness. An accurate account of yesterday’s interaction can be useful for reporting. It is less helpful when someone needs to decide how to handle the customer standing in front of them now.

A different chair at the same table

In 2020, Reltio appointed Chris Hylen chief executive. Sood became chief technology officer, founder, and chairman, with responsibility for product, technology, and security. The change gave him a different set of duties inside the business he had started.

The following year brought a financing milestone: $120 million raised at an approximately $1.7 billion valuation. In their joint announcement, Hylen and Sood connected the occasion to the company’s beginnings and thanked the people who had helped build it. Sood signed as CTO and founder. By August 2022, he was back in the chief executive role.

The changing titles complicate the customary founder portrait, which likes its protagonist to occupy one chair throughout. Reltio’s history leaves room for the technical work and the operating work to demand different arrangements. Sood’s interest in building remained evident. Asked what kept him going, he described the attraction of another difficult problem.

“There is always a new, bigger, better challenge to go after.”

Manish Sood, 2024

AI still needs to know which customer

As enterprise AI became a larger part of Reltio’s work, Sood made a case that connected directly to his earlier career. A system asked to reason about a business needs dependable information about that business. The names must match. Relationships must be intelligible. An agent acting on an outdated supplier record introduces a practical problem well before anyone gets to admire its fluency.

He has also argued for using AI inside the work of data management itself. Matching records can involve laborious checking. Machine learning can help make recommendations, while conversational interfaces can make software easier to use. In his public explanations, AI is both a capability Reltio incorporates and a customer initiative its data is meant to support.

There is a more skeptical side to the argument. On a Nasdaq panel about AI washing, Sood discussed the risks of exaggerated claims and the need for reliable data and transparency. The distinction has practical weight for a buyer: a label on a product says little about whether its output can support a decision.

By his DataDriven26 keynote, the vocabulary included a system of context. The task was to connect entities, relationships, rules, and meaning so agents could deal with business questions spanning several systems. That is a broader ambition than keeping a customer list clean. It also remains recognizably connected to the work he began before the current AI cycle.

Fifteen years, and another beginning

SAP announced its agreement to buy Reltio on March 27, 2026, and completed the acquisition on May 7. The stated purpose included bringing together SAP and non-SAP information for enterprise AI. Sood described the combination as a way to extend Reltio’s work across customers’ mixed technology environments.

September added a different sort of product. Forbes Books published Agentic Intelligence: Strategy at the Speed of Data, which Sood co-authored with Venkat Venkatraman, professor emeritus at Boston University. Its subject moves beyond a software purchase toward how organizations connect information, make decisions, and put intelligent systems to work.

The collaboration fits the direction of his public conversations: data has become a leadership concern as well as a technical one. A company can buy another application relatively quickly. Getting its people and systems to work from a dependable account of the business takes a more sustained effort.

That leaves Sood with work familiar enough to recognize and changed enough to keep him interested. In 2011, the argument concerned whether companies would bring their core information into the cloud. In 2026, more of the argument concerns what people and software should be allowed to do with it. Before either can act sensibly, somebody still has to get the records to agree.