The most critical decisions in the world are driven by people’s understanding of unstructured content. It all comes down to the abilities of subject matter experts to make those decisions entirely driven by whether they’re looking at the right piece of that data at the right time. Is there a solution for these companies to view and process the correct information? Sorcero CEO Dipanwita Das shares ways she discovered data analytics can produce cognitive solutions in life sciences.

Subscribe to the Product Talk podcast on Spotify and Apple Podcasts, and make sure you never miss a single product conversation with leading product executives. New episodes live every Wednesday. 


On Searching Out A Solution for Data in Tech and AI

When Di started the company, she saw a problem that needed a solution. She saw several areas that were stymying certain companies, and it came down to data analytics and the time it took to find the right data. This is how Sorcero was born.

“It’s a marriage of discovering a business problem, first-time, secondhand, thirdhand, persistence for a number of years, meeting the market, and a changing regulatory environment, as well as a certain kind of technology becoming production-grade to enable us to solve this problem. Let’s start with the business problem. First, over a decade of engaging with technical content, STEM content, however, you want to label it across public health, climate change, criminal justice, reform, gender, etc., the most critical decisions in the world are driven by human beings understanding of a bunch of unstructured content. It could be books, papers, articles, social media feeds, and a panoply of other sources. But that subject matter expert’s ability to make that decision is entirely driven by whether they’re looking at the right piece of that data at the right time if it’s been contextualized properly to that design, or that to that particular moment in time where the decision is taken. And last, but certainly not the least, is it aligned with the strategy and the goals of the organization that they are a part of? These three things are absolutely critical, which means that any cognitive solution that’s delivering analytics has to have the customer’s context perfectly encapsulated within the tech in the product itself. So that’s the first thing. That business problem is still quite unsolved. AI, of course, presents us with certain options, and certain ideas and approaches that may actually be able to solve the problem which has been persistent. 

Number two is regulations. Regulations are very exciting, very interesting, and extremely panic-generating across the regulated industries, which are healthcare, life sciences, insurance, and financial services as it pertains to life sciences and healthcare. They’re really facing a very fast-changing regulatory environment as it pertains to how they’re using data to drive decisions, how they’re using data to justify how patients are treated, and making sure that they’re doing a great job of it. That means that they need to collect more data, but much more importantly, they need to know how to deploy it and operationalize it to satisfy the regulatory requirements for them to be able to stay in business as well as to compete. 

The third is transfer learning. This is an incredible new approach that really started becoming production-grade from 2018 onward, and that is very key to how we approach AI. That allowed Sercero the company to start solving this business problem using an analytics platform. So these are the three big whys behind Sorcero. To really put in a single sentence, we are an analytics platform that targets life sciences and healthcare. We specialize in making sense out of scientific content, usually unstructured, although sometimes structured as well, we deliver very complex insights and decision-making in workflow, and our users are almost always MD, PhDs, or someone who has that kind of expertise.”

On Bringing Data Analytics to Life Sciences

There is a need to sift through all of the data for those in the life science industry. Di saw the opportunity to provide analytics for companies in order to have them thrive and compete. 

“In terms of a category strategy, when I started Sercero all the way back in April 2018, I started with customer discovery interviews, so I probably spent between 30 and 45 minutes with about 300 people across seven or eight different industries, all the levels of the hierarchy. The real problem was that they couldn’t find anything, and we would have ended up being an enterprise search company. The real problem was, every time I spoke to a person who worked in a life sciences organization or a healthcare organization, or a financial services organization, they were really struggling to know what parts of their total data set to use to inform their decision. It’s not about making sense of it all. It’s not a one-ring approach to problem-solving. It’s really about specificity and accuracy, and really understanding the user’s environment and context. So that’s one thing. These industries that really need to use AI are also the most risk-averse, they’re highly regulated, and the cost of one wrong decision is extremely high. So their expectations around AI are often a little nuts. They will not forgive AI a single problem or a single mistake, where human beings are no more than 87 to 92% in accuracy. So it also is a very unique space where scaling a cognitive solution is very, very difficult. 

I found that many of these industries are really not ready. It is one thing for us to paint a picture and say if this were true, then their lives would be that. But in examining the ‘if this were true,’ we often find that these industries are not quite ready to start applying AI. They’re still trying to figure out what their data should be. Is it clear mean enough? Where is it stored? Then you go into the regulatory data privacy and all of that nightmare. So that’s one thing about what we found to be true about these industries. 

Let me talk to you about why we chose the market we chose. So we experimented and had successful pilots across a number of different industries, but we chose life sciences because of exactly where they are in their lifecycle. Not only have the life sciences organizations really realized that they need to get a handle on their analytics, but also that they’re being hit by regulations extremely fast. They really need to stay ahead of compliance. Every great business decision, the one article that talks about a new use for that drug, the one article that talks about a combination therapy that’s showing extreme promise in the market …  anything that is going to help them get to that point as fast as possible is a winner. The answer to that is Sorcero. 

What is unique about Sorcero is as follows: Number one, we deliver in workflow analytics and in workflow decision-making to the people who work with content every day. We are not a dashboard that folks look at once in a while. People actually work using Sorcero analytics on a fairly regular basis. Number two, we have the user side and builder side. Our users can just consume the analytics without having to think about what’s under the hood. At the same time, everything we do is transparent and auditable. From the builder’s side, we can also open the hood. They can peek under it, and know that there’s a system that is transparent and they can trust. Number three is in our specialization with unstructured data. We also have built a unified ingestion framework, which allows us to pull in data sources and format-agnostic in a matter of hours where we can add a new source of information. These are three things that make us quite unique and quite differentiated, and allow our customers to not just solve one problem but many on our planet.”

On What It Takes To Be A Product Leader

Product is everything. You probably have heard this a thousand times or more, but it is still just as true as the first time it was said to you. There are some non-negotiables that go into being a great product leader, and Di breaks down what she looks for as a CEO.

“I want to unpack the statement ‘product is everything’ and break it up into two. So from a CEO perspective, I believe that everything a startup does is product because it’s creating value for different stakeholders in wholly different ways. There is a value created for investors for our own team, and for our customers, and for our advisors, and they’re different. So everything Sorcero the company does is a form of product. Then there is our product itself, which is this language intelligence platform where we are marrying certain really smart approaches to AI to bring in accuracy and trust and scale into scientific or science-driven markets. 

When I look at a product leader, and when I look at that, I see a great product leader as one who knows how to feature a product, build a product, define a product-driven by how it is going to actually capture the full market. At the end of the day, I want the product leader at Sorcero to have their minds fully occupied by how is this solving the market problem? How is this feature that I’m pushing for going to get us more market? How is this going to keep us ahead of the competition? How are we completely always leveling up? Those are the big questions that I expect a product leader to keep in mind, which is not too different from the questions I think about: how are we going to be one of the best startups, how are we going to make sure that we have the best in class numbers around our unit economics? Similarly, I would expect the product leader to think that way. 

The product leader should serve as a teacher to our customers by often teaching them how to think about our product beyond the first use case. Our strategy, like every company, is to learn and expand. Let’s say we learn with problem A, and we’re doing a great job and solving the problem. I would want our product leader to start talking to our customers about the next problem, starting to open up this world for them where possibilities they hadn’t thought of before are brought to the table and they’re now actively thinking of them. At the same time, the product leader should be teaching our own team how to come to them with more appropriate requests with better information, how we should present our information to the product team so that they know what to do with it. Give us great frameworks. … I don’t like being told what to do specifically, and I don’t think anyone should be, but I believe in frameworks. If someone gives a really great framework, that’s amazing. 

What are the non-negotiable skills? First and foremost, no matter what the level, one cannot have forgotten how to do work. So they should still be able to roll up their sleeves and knock out a PRD (product requirement document), if required. They should still be able to roll up their sleeves and write up a PR FAQ, they should be able to still get on the phone and do a customer discovery or reach out to someone on LinkedIn. So above all, I value people who enjoy doing the work as much as managing people and organizing teams. Number two has to be an intuitive thinker. We’re never going to have 100% of the information that we would like before we make a decision. So we really, really want people who are able to have enough of a gut in intuitive thinking or enough analogous thinking how they want to describe it, to be able to make the best decision for the company with not all of the information. Number three, they have to be incredibly curious about pretty much everything in the world. This is the other part that I found to be valuable: You don’t know where that right statement, the right argument, the right product inspiration is going to come from. It’s extremely important that they have an insatiable thirst for knowledge, irrespective of what field that knowledge is about, and they are obsessed with finding solutions from everywhere.”

About the host
James Gray Reef Technology, Head of Data Operations

Seasoned executive with three decades of experience envisioning, building, and operating software products and mission-critical platforms at startups to large organizations such as Microsoft. Throughout his career, James has held a diverse set of roles and developed expertise across leadership, product management, data science, IT operations, consulting, software engineering, and sales. As Chief Product Officer at Products That Count, he leads an online platform to help organizations learn the craft of product management and win as a market leader. As a coach, he leads mastermind circles to facilitate peer-to-peer learning across product management executives. As a podcast host, he interviews CEOs of companies held within Mighty Capital’s portfolio. Creator of the Career Strategy Framework and an online platform that teaches and coaches professionals how to reach their full career potential. Learn more at https://www.jamesgray.io.

Leave a Reply