Pune, Bangalore, Mumbai and Delhi have become synonymous with being tech hubs in India. What if someone were to tell you the future of Indian tech is situated in second-tier cities? In a conversation with Findability Sciences, Founder and CEO Anand Mahurkar spoke about the organisation’s base in Aurangabad and the success of moving away from metropolitan cities. With its roots in Boston and India, Findability Sciences is an Enterprise AI company helping businesses realise the potential of data and become data superpowers. Analytics India Magazine spoke with Anand Mahurkar to know more about the company’s offerings and plans for India.
AIM: What problem does Findability Sciences solve?
Findability Sciences is an enterprise AI company. We have three technologies in our AI enterprise offerings; machine learning, natural language processing and computer vision. We help businesses answer ‘what will happen’ and ‘what to do’ in their businesses. Reports and analytics generally explain the past, but our AI tells businesses what will happen next. Findability sciences solves the digital transformation challenges for companies.
When I founded Findability Sciences, I was looking for technology with the ability to find information, hence find-ability. As we continued to build this technology, it became the basis for any AI solution. To teach AI, we needed history and data. Our initial platform provided the ability to find information by connecting internal data to external data. Think of it as a Google-like interface for your internal organisation. Later on, since we already had cracked the data problem, our next level of the stack was adding AI technologies and solutions.
AIM: What are the various solutions Findability Sciences has built to help traditional Indian enterprises unlock their data potential?
Most companies today cant find information internally easily. The workforce spends a huge marker of time finding information in most organisations, but we can solve that. Everybody wants digital transformation today. It begins with data and ends with AI, and our full-stack provides these solutions.
We have four offerings. Findability.ai is our core intellectual property that we offer in terms of licence for various applications. Secondly, our data science labs, Findability.dsl, is a collaboration between our team and customer teams to build intellectual property for them. Most companies want to build their own AI for competitive advantage; we help with that. Our third offering is Findability.inside, which provides a unique solution to power legacy technology. We can embed AI into year old solutions. It also includes powering legacy hardware like a scanner to scan a document and output the key points/summary of the content. This will be integral to India, given the legacy hardware and software present here. Our final solution is Findability.labs. We provide related data solutions to legacy organisations without data platforms.
AIM: Can you share some case studies?
We are working with an advertisement technology company for portable headphones with 13-14-year-old software. They don’t have predictive analytics in the advertisement technology. Apart from location, they dont have any targeting indicators. With our AI, they have better-targeted advertisements. Another solution we are working on is one of India’s largest secondhand used car sale companies. They use our predictive analytics to predict the prices on the platform. AI has developed car prices based on historical data and price predictions. We work with one of the largest immigration consultants in India, and we have powered their platform with natural language processing. Their conversational computing platform answers repetitive, criss-cross, and research-based questions immigrants may seek. Recently, we have collaborated with Aviva Insurance, where you can do all your insurance transactions through conversational computing.
AIM: Could you elaborate on the tech stack powering your platform?
Our tech stack is divided into four layers. First is a collection. We collect data from internal, external, structured and unstructured sources. For example, suppose I am a manufacturing unit with an SAP as an ERP system. That’s my internal ERP data, CRM, financial systems, IoT sensors and more. Now, I can have external data in terms of weather, social-economic situations, geopolitical situations, or news. These are structured content beyond the firewalls of my company. Our technology stack collects the data in a central location and unifies it. So the second part of our tech stack is unification. Our algorithms at unification do entity resolution, data matching and more.
Once the data is unified, we move on to the processing layer. This layer leverages our three major algorithms discussed before – predictive and forecasting, NLP for chatbots and computer vision to process media. Post-processing the data, our fourth layer is presented. Presenting differs according to the organisation, sector and application; there is no standard format. For instance, if our AI indicates possible machine failure in the manufacturing industry, the presentation layer can be an SMS to the factory manager noting this machine is likely to fail. Alternatively, there can be fancy dashboards or data going back into the legacy system like SAP, depending on the use-case.
Python and Java are our programming languages on the front end. Our proprietary connectors connect various data sources and connect to unstructured data.
AIM: What is Findability Sciences’ GTM strategy for India?
We have technology partnerships, system integration partnerships and customer partnerships. First, we have a partnership with IBM; it is a global partnership. We have integrated our technology with IBM, especially IBM’s data-related technology, particularly that latest platform is called ICP for D, which is IBM cloud for data. We have integrated our predictive engine with that, and any IBM users can use it. The second partnership is with Snowflake. Snowflake is another data warehousing and data lake company where we have integrated our technology. We also have system integration partnerships with Tech Mahindra, ITC and Infotech.
AIM: What is the state of digital transformation in India?
In the past year, post-2020, the state of digital transformation has been growing very rapidly. The statistics say that India will touch a billion-dollar transformation by 2023. When we use digital transformation, it is used very loosely. We use AI-powered digital transformation. We are talking about automating and modifying current processes. AI-led digital transformation will be about $800- $900 million spent by 2023/24. So roughly a billion-dollar expenditure will happen to practically every legacy company. In India, resources are fairly cost-effective, but the best thing that has happened is the skill of data science. It needs three skills: statistics, mathematics, computer programming and domain knowledge.
AIM: How can we enhance digital transformation in India?
Most companies have domain knowledge, but external help is required to customise solutions. It needs this combination of stat math, computer science, and domain knowledge. We bring the mathematics and statistics combination to the organisation’s domain knowledge.
AIM: Tell us about your strategy of operating from secondary cities.
I have a huge focus on supporting secondary cities in India. We have our operations centres in Aurangabad, Maharashtra – ours being the only software company in the town. We want to take off the loads from metro cities. Findability is headquartered in the US, but I wanted to contribute to India. I disagree with the model of only loading the Metro cities. It is not sustainable. We are already seeing the crippling infrastructure, the cost, the travelling time and the difficulty of getting a job. Secondly, I had studied engineering in Aurangabad. Our office in Aurangabad supports 70-80 families with a world-class infrastructure and above industry level salaries, giving us a less attrition rate and world-class projects. We are creating quality employment.
AIM: In this context, what is your future strategy for India?
Globally, there is a shortage of data scientists. India offers a good base of mathematics and science students and will continue to become the hub part of global data science activities. We want to double our account in India. We also want to explore other secondary towns like Belgaum in Karnataka. We want to expand and create world-class infrastructure, create quality jobs in AI and hire more people in such towns.
AIM: How can India encourage this kind of growth?
The data science activity does need a white-collar job or blue-collar job people. It needs a new collar job for people. IIT engineers are great, but they may not all be ready for data science activities. They need training. The industry calls this era for technologies a cognitive era. Traditional schools are not going to support the real requirements of the job. You need training; you need to groom these people. We have partnered with local colleges in Aurangabad and Boston, where we take talent right out of the college. We train them in statistics, mathematics, data science, etc. No university today can produce a ready-made data scientist. Therefore, whether it is in a primary town or a secondary town, if you have a base talent, AI organisations like us need to take responsibility for training and creating these new collar jobs. Our attrition rate is less than 10%, with people working with us for 6-7 years. This strategy is proving to be successful. I, myself, groomed my career and success from Aurangabad.
