Artificial Intelligence is the new rage. Global tech firms are deploying their best minds to constantly evolve machine learning systems and create more use cases. At the same time, there has been a flurry of startups which have picked up different segments & market opportunities to leverage this glorified technology. Not to forget, the big fish are also acquiring smaller AI firms wherever they smell innovation.
Now talking about why this field is gaining so much attention. Automation has always been a requirement for all types of businesses, and with machine learning tech in picture, things are just going to get more and more efficient. Corporates look for automated solutions to cut down on manpower costs, increase process efficiency and build robust systems that can last for years. Yes, technology is usually the answer for covering all these requirements, and an evolving autonomous tech takes the solution to the next level. And since being digital is all about knowing your customer, the tech offers hyper-personalisation too.
Chatbots are finding use cases in different consumer facing applications. From simple rule based bots that answer to pre-defined questions, to evolved ones with NLP & AI capabilities, we are witnessing all kinds of automated assistants coming up. Understanding that businesses can leverage this tech beyond just a simple chat interface, Bangalore-based Racetrack.ai has taken up the job to make sales & customer support much easier for companies across sectors using its AI helper Marvin.
Racetrack.ai is an artificial intelligence driven platform that enables sales support for enterprises, and also addresses real time customer issues. The startup was founded in October 2015 by Subrat Parida, Co-Founder & CEO and Navneet Gupta, Co-Founder & CTO, with a vision to empower businesses to increase customer acquisition and retention by enhancing Customer experience using deep learning.
The 27-membered team is on a mission to increase its clients’ revenues by enhancing human capabilities using AI. It has built different sales enablement applications, including communication bot ‘Marvin’, that can communicate with end users across chat, email, audio and video.
The firm is working with 9 clients belonging to different sectors, namely – Education, Healthcare, Real Estate, Investment Banking and eCommerce. Subrat shares that for each sector they do preliminary customisation depending upon the need, and post implementation at the client’s end, the bot possesses self learning capabilities that allows it to become smarter.
What Does Marvin Do?
Marvin sits at the entry point for all customer interactions – be it the client’s app, website or support desk. It has mainly been programmed to drive sales conversions by responding to customer queries. For instance, as a replacement of live chat on eCommerce websites, it can suggest a visitor what to buy on the basis of their preference. Moreover, it uses proprietary algorithms to learn from previous chats which can be used for further optimisation.
Such conversational bots find a good use case in customer engagement and support, where they can be trained to assist users with their questions.
Subrat takes pride in Marvin’s NLP capabilities, and emphasises that “It’s not a chatbot, it’s a communication bot that operates across 4 major communication channels – email, text, audio and video”. Prospects & customers of different firms have different preferences for reach out, hence they have built audio NLP capabilities as well. For instance, according to their data, most queries for their real estate clients come via telephone.
We enquired how often do visitors really need to chat to a human instead of a bot, to which Subrat replied ‘In a very few instances, if at all requested, then Marvin patches the prospect up with a representative via chat, call or video.’
Since the whole goal of such a system is sales enablement, Marvin, comes with a sales dashboard that gathers insights on performance data on conversions, user sentiments etc. In case of audio or video queries, it records the conversation and send a copy to both the customer and the support team.
Building A Sales Focussed AI Driven Platform
One of the biggest roadblocks in building and growing Racetrack.ai was the lack of trained resources in the field of machine learning. The founders overcame it by hiring PhD backed professionals holding expertise in the fields of linguistics, machine learning, and deep learning. Who helped in building the product with TensorFlow libraries, with Python, Java, Node.js, Angular.js, MongoDB and MySQL in its stack. They also hired enthusiastic professionals and trained them to become a growing AI firm.
With all this in place, the founders still found lack of business data to train the system an obstacle. ‘Staying updated & matching with the Human Intelligence’ was another challenge according to the founders. ‘We made strategic business & research tie-ups to scale faster, and signed NDA contracts with customers to get access into their real time data.’ shares Subrat.
It’s working closely with IISc Bangalore to help them with research to train the system.
Moreover, the challenges weren’t restricted to building the product. Since AI is still in its nascent stages of penetration, making enterprises adopt an AI platform isn’t easy. The team is nonetheless making efforts to educate its prospects about the benefits of this technology, and eliminate the notion that ‘AI will replace human employees’.
Growth & Numbers
As mentioned above, the year old company has 9 paying clients at present. On an average, the system interacts with over 11000 customers in a month, processing over 70000 queries.
Racetrack.ai was bootstrapped from October 2015 till September 2016 and raised an undisclosed amount of investment in angel round in September 2016.
Now the immediate plans of the firm is to expand globally. It’s in discussion with larger brands and customers from UK & US. After analysing the response till date, it plans to focus majorly on the healthcare sector.
About the Founders
Its CEO Subrat Parida is a Business Intelligence expert who is a two time co-founder of successful startups with one rationally acquired and another running successfully. He is an IIM Bangalore alumni, and believes in empowering employees to transform them as future leaders, at racetrack.ai he realises his dream of “We together” concept work.
CTO Navneet Gupta’s passion lies in interactive AI driven products. He is a BITS Pilani graduate and a technology enthusiast. He brings in rich learnings from product R&D departments of Infosys Ltd. And Edgeverve Systems. He truly believes in “No idea is stupid”.