(May 4, 2024) In today’s data-driven world, companies collect vast amounts of information, hoping to unlock valuable insights. Those insights find their way into reports and presentations but often end up not being used, for a number of reasons. For instance, the reports could be too complex and full of jargon for non-technical people to translate into actionable steps, or have so many potential improvements that teams become overwhelmed and don’t know where to focus. Change can also be difficult and organisations could struggle to adapt their processes or culture to these new insights, even if they might be beneficial to them. This is the gap that Shub Bhowmick and the team at Tredence are hoping to bridge.
For over two decades, Shub Bhowmick has grown as a ‘problem solver, entrepreneur and technology leader’. In 2013, he co-founded Tredence, a data science and AI engineering company that now has over 1000 employees with offices in Foster City, Chicago, London, Toronto and Bengaluru. Their clients include over 30 Fortune 500 companies in a wide range of sectors, as they work specifically towards solving this last mile problem in analytics.
Shub Bhowmick
Discovering the potential in data analytics
After graduating with a B.Tech degree in Chemical Engineering from IIT-BHU, Shub went on to do an MBA at Northwestern University’s Kellogg School of Management. He has held high ranking positions at Diamond Consultants (currently PwC), Mu Sigma, Liberty Advisor Group and Infosys. His career in data analytics began at a consultancy in Chicago, Diamond Consultants. “Around 12 or 13 years back I was involved in a data analytics practice that Diamond had created, and I was deeply influenced by this experience,” he recalls, in an interview with Nasscom. “I realised for the first time how significantly data management can actually have an impact.”
From there, he moved to Mu Sigma, which he describes as another deeply inspirational experience. “I was able to see how data analytics services are not just an add-on service offering – at Mu Sigma, it was the core and basically the only service offering,” he said. Over a decade ago, Mu Sigma was at the forefront of last mile services in data analytics, and worked to provide actionable business insights for Fortune 500 companies. They offered a range of data analytics services, helping clients collect, clean, analyze and interpret vast amounts of data. They even had their own Art of Problem Solving platform, which is a set of tools designed to help clients translate vast amounts of data into concrete solutions, emphasizing actionable strategies over reports. “They were able to create a very interesting business model around this,” Shub recalls. “The industry was also starting to attract some really good talent and nurture them into future leaders.”
Still, Shub would wonder if there was a better, more efficient way to deliver this service. “Back then, in the analytics industry, most of the companies and providers relied on a manual process, on a value chain that involved pulling data typically into a throwaway MS Office based data file, using Excel for the analysis, some bit of SaaS, Power Point based summary and delivery.” This was a decade ago, when cloud technology was still very new and not really in use. “It was basically a sales automation platform, there was no Azure or Google Cloud,” Shub explains.
The tech-centric approach to business insights
Around 2012-13, Shub Bhowmick, then based in Silicon Valley, saw an opportunity to start a different kind of data service. “We brought the essence of business analytics, which is the focus on business insights, but we combined that with engineering, to deliver insights in a more efficient, value-driven, adoption-focussed way.” This set them up from the competition too, as they moved away from business analysts and the manual processes towards a more technology centric approach. “We were using big tech to automate portions of the value chain and create greater scale and speed in insights delivery.”
Over the last seven years or so, cloud-based platforms and hyperscalers took the world by storm, and production workloads quickly moved to cloud tech. Tredence Inc didn’t waste time in getting on board, and began developing ‘cloud-centric capabilities to deliver analytics services with a focus on adoption and shortening the time to impact,” the Global Indian explains. “Cloud and data analytics are very much intertwined and will continue to be so in my opinion, as enterprises invest in cloud native AI capabilities,” he adds.
The age of generative AI
The arrival of generative language model Chat GPT was another game changer. “I had never heard the use of the word ‘hallucination’ in my industry until we all came across this explosion created by OpenAI’, Shub said in an interview. “Since then it’s been the only topic everybody’s talking about, especially in technology.” He has watched the ecosystem grow, from a time in Silicon Valley where companies hired AI experts to work in isolated corners of office buildings, to now, where titles like ‘Chief AI officer’ are common, and AI developers are a core arm of big tech.
“We had talked about AI for a long time, we used to call it advanced data science and applied analytics, and just AI for the longest time. Now we call it Generative AI but the idea is not very different,” he says. “It’s about how you take data, information that you already have within your firewall, or leverage other data sources and then help your executives make more meaningful decisions to improve their business.” At Tredence Inc, he says, the team is working on fine tuning foundational models, and prompt engineering systems to cater to their existing clients, and provide them with a wide range of highly customised insights through AI language models. Coding assistance is another important segment, as the industry begins to recognize that generative AI can significantly improve the productivity of all kinds of engineers.
In March 2024, Tredence decided to invest 10 percent of its annual revenues in developing GenAi and advanced AI capabilities across engineering, customer experience, machine learning operations, supply chain and other verticals in data analytics. Through this the San-Jose based startup is looking to grow revenues by 40 to 50 percent, Shub Bhowmick told ET. “We’re building AI language models by fine-tuning foundational models. These models need not be large in size, we are using public and proprietary data of our customers to create agents to serve their unique needs in sectors such as retail, consumer goods, healthcare, telecom, banking and financial services and manufacturing.”
