Venkat Srinivasan and the six-binder problem

At Apple in the late 1980s, Venkat Srinivasan encountered six massive binders of plans for improving business processes. He was designing a knowledge-based system for the company’s Global Financial Services group. The binders described how work ought to happen. When he asked a senior executive about putting the plans into practice, the answer involved ten years and still finishing only a small fraction.

In Srinivasan’s account, the binders never became working processes. The division changed shape as the business changed around it. There is a particular corporate comedy here: an organization can finish explaining its future just as that future becomes obsolete. The filing system enjoys excellent job security.

The episode appears near the beginning of his book, The Intelligent Enterprise in the Era of Big Data. It also offers a useful way into his career. Again and again, he has worked on the distance between knowing what to do and making a machine do it. Credit decisions, business documents, reading comprehension: different settings, a recurring engineering question.

The accountant at the computer center

Srinivasan grew up in New Delhi and studied commerce at Delhi University. He qualified as a chartered accountant through the Institute of Chartered Accountants of India. Finance was his formal route into work; computing became the interest that kept pulling him sideways. That combination would eventually matter more than either discipline on its own.

His move to the United States began with a choice between a possible partnership at Deloitte Haskins and Sells and doctoral study. He chose Cincinnati, borrowing money from a friend for the airfare. While pursuing his doctorate, he spent time around the computer center despite being, by training, the finance and accounting person. The detour was becoming a direction.

He later joined Northeastern University’s business faculty. Research appealed to him, and consulting brought him into contact with companies including IBM and Apple. His account of the transition to entrepreneurship is practical: clients needed software they could keep using. A consulting assignment could identify a problem; a product could remain behind to address it.

A credit decision becomes a software problem

Before RAGE, there was eCredit.com. Srinivasan served as its founder, chairman and CEO. Its territory was credit, financing and receivables management, spanning business-to-business and consumer commerce. These are fields where policy becomes consequential in a hurry: someone must decide whether to extend credit, on what terms, and with what supporting information.

His expertise crossed knowledge-based systems, computational linguistics and business process automation. In ordinary language, that means trying to make business knowledge usable by software. An accountant’s understanding of a financial statement and a programmer’s representation of that statement have to meet somewhere. Much of the interesting work takes place at the meeting point.

Northeastern now lists him as founder and managing director of Innospark Ventures, and as a former founding chairman and CEO of RAGE. It also records his service on the Center for Emerging Markets board since 2012. His relationship with the university continued after he left the faculty. Entrepreneurship gave his research a different working address.

The business must be allowed to change

By 2009, Srinivasan was arguing that moving work to a cheaper location would eventually reach its limits. Wage differences could narrow. More fundamentally, relocating a cumbersome process did little to make that process adaptable. His argument for RAGE was that companies needed software capable of accommodating change without starting another lengthy development cycle.

He described reusable, adaptable components: business rules, decision trees, computational expressions and language processing. The ambition was to let a company represent the work it wanted done while reducing the translation between business intent and executable software. A policy manual could explain a procedure; a software framework might help carry it out.

“Fundamentally, I am a creator,” he said in 2009. The description joins the professor to the founder rather neatly. Research and product development both begin with an unsatisfactory answer. A business then adds an awkward requirement of its own: someone has to want the improved answer enough to use it.

Genpact had already tried the machinery

On March 14, 2017, Genpact announced an agreement to acquire RAGE Frameworks. The announcement followed an 18-month strategic partnership. The companies had already worked together on customer engagements, including management reporting and financial spreading, the work of turning financial statements into structured information for analysis.

The platform’s applications included reading financial data, contracts, news and business reports. RAGE combined a no-code development platform with expertise in AI, automation and financial services. Its headquarters were in Dedham, Massachusetts, with operations centers in Pune and Belgaum, India. The business was international in its working arrangements as well as its ambitions.

For Srinivasan, the sale also marked a personal boundary. In 2017 he explained that he had promised his family he would leave operating roles when he turned 60. He wanted to return to incubating technology and social enterprises. An acquisition usually arrives in a founder’s story as a financial event. Here it also served as a way to rearrange his working life.

Three requirements, one unfinished argument

His book put the argument into an organizational framework. Wiley lists its publication in October 2016. The central combination is efficiency, flexibility and intelligence: an enterprise needs to execute work, adjust to changing requirements and interpret developments inside and outside the business. Improving one dimension while neglecting the others leaves familiar frustrations intact.

The subject matter is deliberately close to everyday corporate work. Natural language processing belongs alongside business transactions. Time to market belongs alongside systems architecture. Case studies sit beside theory. For a former finance professor who became a software founder, a book that moves between concepts and implementation is a fairly faithful choice of form.

The six binders make sense in this framework. Their contents might be sensible, yet their usefulness depends on whether the organization can put them to work before circumstances change. A beautifully documented plan has a timing problem. Srinivasan’s book treats the technology underneath the business as part of that problem and part of its possible solution.

The next document is a schoolbook

EnglishHelper gives his interest in language a different destination. Founded by Srinivasan and led by global CEO Sanjay Gupta, the company develops technology for English proficiency. Its RightToRead initiative brought ReadToMe, a reading and comprehension product, into thousands of public schools in India. The company also describes a footprint extending into Africa and Latin America.

The question becomes whether technology can help teaching and learning within the setting that already exists. That is a demanding test. A classroom comes with teachers, timetables, textbooks and practical constraints. Any proposed tool has to find a place among them. The quality of a demonstration alone cannot settle whether it belongs there.

Srinivasan and Hemavathi Murthy co-authored a 2021 research paper examining an AI-based multisensory literacy intervention in Indian government schools. It focused on English reading and comprehension. The paper brings him back to the academic practice of testing an intervention and examining outcomes. The founder and the researcher still share a desk, at least intellectually.

An answer should leave a trail

His later projects include Gyan, where he is listed as founder and chair, and PrismX, whose focus is automated software development. Gyan’s team includes CEO Joy Dasgupta, whose earlier experience spans RAGE and Genpact. The names change across ventures, but there is continuity in both the people and the problems.

Gyan announced a language model and research engine in March 2023, emphasizing explainability and the ability to trace results. Srinivasan’s own writing makes the broader case for combining expert knowledge with learning from data. He argues that context, traceability and the ability to work with limited data deserve attention alongside a model’s apparent fluency.

These are his technical positions, and they help explain his choice of work. A system used for a business decision must be open to examination. If the answer changes, someone may need to understand why. The attraction of a machine that explains itself becomes clearer when the audience includes people responsible for the consequences of its output.

“Fundamentally, I am a creator.”

Venkat Srinivasan · 2009

A different seat, familiar questions

Innospark Ventures, launched in 2018, gave Srinivasan a way to support other founders. Its stated strategy targets pre-seed and seed companies whose technical founders bring domain expertise to applied AI and deep technology. The firm offers help with AI strategy, fundraising and business decisions. His present role is founder and managing director.

In December 2024, he described a portfolio of roughly 30 companies and an interest in early ventures with substantive intellectual property. The portfolio included Phaseshift, working on materials discovery. He also explained that Innospark began with a Boston focus, then responded to opportunities brought by teams farther afield. Those figures describe that interview’s moment, rather than a permanent company tally.

His public life also includes board service at New Profit. At an American India Foundation New England gala, he and his wife Pratima served among the chairs of a fundraising effort that raised $1.03 million. The connections run through universities, companies and civic organizations, each with its own version of the question of how useful work reaches more people.

Room for coffee

The quieter detail in his PrismX biography is that he enjoys exploring Boston’s coffee shops and spending time with his wife and two daughters. It supplies a welcome change of scale after enterprise architecture. Some decisions can be made without a decision tree, although a café menu may occasionally test that proposition.

His career has moved between studying organizations, building products for them and financing people who want to build the next ones. The work carries an old question into new settings: how much of what people understand can be made useful in software? The binders at Apple belonged to a particular corporate moment. The difficulty they represented has kept him occupied for decades.