We had the chance to sit down with founder and CEO of Intellizence Sachi Komarasamy to discuss the nuances of article intelligence (AI) and human intelligence. This article was adapted from a podcast interview hosted by LAC’s Chief Business Development Officer Mario Thériault. The full audio recording is also available to listen to below.
The start of Intellizence
Mario: Sachi, could you tell us how you decided to start this platform for market intelligence using AI or machine learning? What was behind your thinking when you launched Intellizence?
Sachi: I was in business development in the technology industry, so there were certain instances where I won opportunities, or lost opportunities because of availability of the timely intelligence, as well as the lack of intelligence. What I have found is a lot of this information about my customers was already in a public domain, but getting those right signals at the right time was a challenge. To try to solve my own problem, I ended up building the solutions. I found that it’s an interest not only from sales people, but from a variety of people across the sectors.
Mario: What year did you launch your platform?
Sachi: In 2016.
Mario: Above and beyond what you just shared with us, that you were looking for a solution, and you thought that the content was already in the public domain, did you do some sort of market assessment or more formal business plan? Or did you just follow your gut and say, “I’m going to do this?”
Sachi: I would say it’s a combination of both. I tried some of the interesting solutions in the market, but somehow I was not happy. And then there was a gut feeling because I pitched this with different colleagues, my other friends, and then I found there is a wide space to be at this. So, I started this.
Mario: There’s a lot of jargon around AI and machine learning and market intelligence and competitive intelligence. Why were you not happy with the technologies? What was the gap that you had identified? What did you think was missing that you could build a better mousetrap?
Sachi: I found there weren’t really good solutions specifically built for this purpose to monitor the companies of my interests. Of course, there are solutions like Google Alert, which anyone can use, but it’s primarily like a text match, which generates too much noise. And again, it’s limited to certain sources.
Mario: Could you elaborate on that one point?
Sachi: Let’s say you have around 1,000 customers. Now, the world would like to know, “Hey, I want to know how many of our customers appeared in the last six months?” or “Which customers expanded into Europe?” Or it could be “Which customers are doing a layoff?” These are the typical questions which are asked by the executives in an organization. But can you do this in Google Alert or can you do this in Google Search? The answer is no, because the information is spread across in multiple places, just beyond use. This information, so let’s say, the expansion signal could be an earnings call transcript, or the layoff data could be in one filing. How do you get this information? Most importantly, how do you curate this? These are the problems that we are trying to solve.
Evolution of AI technology
Mario: How has that technology evolved since you’ve started Intellizence over the last four or five years? What are you now able to do that you weren’t able to do in 2015-2016?
Sachi: We started Intellizence with a goal of maybe monitoring a set of companies for an individual. But then over a period of time, we realized people would not only monitor companies, but also the industry trends on the topics. That’s sort of an evolution. The second one is like they would monitor say, thousands of companies. Not just ten. So, now it can’t be humanly solved. That is where we ended up building a lot of natural language processing, machine learning techniques, to not only monitor those companies, but filter out all that noise making it relevant. That is sort of the second evolution, how to scale this platform.
Mario: What’s an example of natural language?
Sachi: That’s language processing. These are certain technologies broadly and what you call, artificial intelligence. Let’s say you are interested in monitoring Scotiabank. A simple search and Google Alert would throw up many news that matches with Scotiabank, which would include ‘Scotiabank stadium’ or ‘Scotiabank tower’. How do you filter out and identify if this is news about Scotiabank, not about Scotiabank tower? There are certain technologies which exist, which would identify these without any human intervention.
Intersection between AI and human intelligence
Mario: Before the advent of AI and these technologies, we were using aggregators, alerts, and people. We still use people because we’ve concluded in the marketplace that there are some things that machines or computers can do, and yet there are still some things that they cannot do. How do you view this continuum of AI and human intelligence?
Sachi: Certainly, humans would always be in the loop. What AI brings in, in terms of the ability to process huge volumes of data at a scale and at a speed, will still need people to identify and look for information. What is the noise or what is the signal? Someone has to train the machine.
Mario: Can the machine or software, effectively say what is noise and what is a signal?
Sachi: There is an element of context there, so only the humans understand the particular context. Someone has to train the machine — this is of relevance, this is noise — the training has to be done by the human, or the analyst.
Mario: Can you give us a real-world example of this? Even if you’re building a software and an AI platform, are you using analysts to fine-tune your searches or your algorithms to weed out and double-check the final outputs? Where is this intersection between AI and human intelligence?
Sachi: We have a team of curators who help to train the machines. Let’s say, if there is acquisition news, it can be mentioned in different ways. It could be like “Company A acquires Company B.” “Company A combines with Company B” or it could be just “Company A plus Company B.” As a human, we can understand “Company A plus Company B” is contextually an acquisition. This is a new scenario where someone has to train the machine. That is where the human curators are coming into play.
Uses for AI in MI & CI
Mario: Back to your market more specifically, would you say that there is a proliferation of AI platforms for market or competitive intelligence?
Sachi: There are different platforms that exist to solve different problems. There are certain platforms focused on certain industries or there are certain platforms focused on certain use cases — a company doing intelligence. Each one is trying to solve a different problem from a different angle. In my view, given the huge proliferation of data and each industry is looking for relevance, I think there is an opportunity for all the players.
Mario: Who would be the users of Intellizence? Is it more strategy people, business development people, a CIO? Who would be interested in such tools inside an organization?
Sachi: We have broadly two customer segments. One is the functions that are responsible for creating the intelligence from the external sources in an enterprise or corporation. The title might be market intelligence innovation or strategy. Another segment is other technology platforms like a CRM or a customer service management who are interested in getting this intelligence and flow-through as part of their platform — primarily for the attention of their customers, their platform customers. Here our customers would typically be the product management team or the data management team.
Mario: Would you say what you’re building and what you’re trying to deliver is easier to convey now? Or do you feel that there is still a phase of educating the prospects and customers about the uses and benefits of such a platform?
Sachi: I think there is a level of education, but the customers are starting to understand the power of such platforms. Even now we are working with a very large global customer, where their management has told them, “You have two people, but now go and get some AI platform and process all this data and give me intelligence for this.” So, it has already started happening.
AI and human intelligence working together
Mario: Is this a new market development over the last 12-24 months? Or how do you see that market evolving?
Sachi: Yes, certainly in the early days, our focus was more about outbound — reaching out to customers, talking about it. But in the last 12 months, especially after COVID, we are putting a lot of output and telling our story on the marketing side. There are people sitting with their problems, but they do not know that a product like ours exists. Now in the last 12 months, we have started getting a lot more inbound requests.
Mario: Where do you think this is going? Ultimately, how will AI and human intelligence still coexist in a similar way, or will that change?
Sachi: I think the future would be more of a collaborative or what they call a hybrid intelligence. There is always a place for human intelligence. It cannot take place with AI. But at the same time, AI brings in certain advantages — whether it is in terms of the processing and speed. There is always a coexistence that would happen. I think we are increasingly seeing our customers spend more time on analyzing this data and answering the so-what questions, rather than spending time on doing the data search, which would be taken care of by the AI.
Mario: On that note, the optimistic view that AI and human intelligence will just keep on growing their collaboration and I share your view that we’re heading towards hybrid solutions any which way we slice it. So, I wish you all the best. Thank you for your time and insights today.
Intellizence is an AI-powered business signal intelligence platform that enables its clients to discover information in target companies as well as stay informed about emerging industry trends and regulatory changes.