Startup Mantra: Giving firms the power of ‘AI workers’

Picture this: The HR department of an enterprise is grappling with heaps of paperwork and deadlines are looming. A familiar scenario in your company? But what if there was a way to break free from the limitations of traditional methods and unlock the full potential of your HR operations? That is where the power of AI automation in HR comes in. Businesses can streamline critical functions such as recruitment, employee onboarding, performance management, payroll processing, compliance reporting, and more. And this is only one among many use cases of “AI workers” across different sectors.

Pune-based startup Light Information Systems launched E42 (formerly known as NLP-Bots), a natural language processing (NLP) based platform to build such AI workers. Founded by Animesh Samuel and Sanjeev Menon, the startup is providing end-to-end automation, Platform-as-a-Service (PaaS), to enterprises ranging from Fortune 500 companies to SMEs and startups across sectors: from automobile majors and pharmaceutical giants, to BFSI (banking, financial services and insurance), telecom and new-age unicorns. By maximising efficiencies and scalability and minimising complexities and costs, E42 is disrupting the enterprise automation space to deliver seamless user-experience.

In the beginning…

Animesh and Sanjeev completed engineering from Kolhapur University in 1996-97. Later, Sanjeev pursued his education in Mumbai, worked with Tata Institute of Fundamental Research (TIFR), and became familiar with NLP. He went to Florida for his postgraduation, where he gained professional experience and even started his first successful startup. After gaining many global customers, he moved his operations to India. He exited the company in 2007-08. Meanwhile, Animesh was working with a couple of companies and then went to the Middle East to establish his company. There, he built a product that matched behavioural compatibility to job profiles and provided soft skills training to companies. He eventually, moved back to India.

Animesh said, “We both reconnected after 16-17 years and formed a company called ‘Light Information Systems’ with the primary goal of doing something creative. Although we were unsure what that was, we wanted to create something that was next generation. Our starting point was the idea that language was the key to human intelligence, and with language, we could exchange data and become more powerful. Similarly, we believed that machines would also need language to become intelligent. We set out to teach machines how to comprehend language, knowing that it was one of the most significant things that needed to be accomplished.

“Sanjeev had some idea about it and we knew that it was one of the biggest things that needs to be cracked. IBM was already using NLP and we believed that with AI we could do a better job than them, but we lacked access to the servers and technology needed to make it happen. Fortunately, Microsoft later helped us by providing $20,000 worth of Nvidia servers and its GPUs (graphics processing units). So, we could try out our algorithms and initial AI models. That is how the journey began.”

Answering machine

Animesh and his team had launched a service called “Poonchh” meaning ask in Hindi. It was a conversational AI service way back in 2014. Recalling his initial success and failures, Animesh said, “For natural language conversations we required data and hence we launched an answering machine hack called ‘Poonchh’ on SMS service. People could ask a question via SMS and receive an answer, and the AI was trained with each interaction. We also launched a system of experts to help us answer difficult questions. This hack became viral in Pune, Kerala, and the U.S., providing millions of questions and answers that were critical for our AI modelling. Unfortunately, in a bid to curtail the spam menace, the telecom regulatory authority increased charges for bulk SMSes from 1 paisa to almost 60 paise per SMS, making it unfeasible for us to continue the hack.

“We then tried launching the service on WhatsApp, but our bot, software app programmed to execute specific tasks, was blocked. We tried switching numbers, but it was not a feasible solution. We eventually tried Telegram, which became the first platform for bots to answer human questions. Although it was an exciting experience, we could not sustain it as a business plan.”

Accelerator

“E42 was selected among the top 11 startups in the GenNext startup accelerator launched by Microsoft and Reliance. Those were exciting times,” said Animesh recalling their hustle between Pune and Mumbai.

“We had an opportunity to present at the Global Silicon Valley Summit in 2016 and then we had said that ‘answering is the next big thing’. We thought they were making billions out of the search business and why would they change what is working for them. But to our surprise, some of them were already working on similar projects like our answering machine. We understood that if they are pursuing it, and considering the data game, they would beat us easily. So, we decided to expedite our work and pivot our business model. We had good algorithms and hence we decided to switch our focus to enterprises,” Animesh said.

B2B model

After successfully raising funds for the first time, Animesh and Sanjeev decided to venture into the B2B space and launched the first version of E42 (formerly known as NLP Bots in 2018).

Sanjeev said, “Today, our platform serves as a one-stop-shop for building AI workers. While most people are talking about chatbots, we have taken it a step further by enabling enterprises to build various kinds of AI workers. It is estimated that 50 per cent of enterprise workers will be AI-driven in the future, so we built a platform to facilitate this transition.

“After our launch, we gained traction with tech-savvy companies in the telecom and automotive industries. All the major Cloud providers also became our partners. However, we realised that being just a bots-solution company would not allow us to scale effectively. Therefore, we pivoted and transformed E42 into a product, which we launched one and half years ago. It is now a no-code platform that enables our partners and enterprises to build AI workers easily. We made it low-code so that anyone can use it by simply looking, dragging, and dropping.”

AI marketplace

Animesh claims that they created the first-ever AI marketplace in the world where one can hire AI workers with just a click. “We soft-launched it initially for our partners to upload and describe their AI workers, but it will be fully functional in another 4 to 5 years. The ultimate goal is for our company to become like the popular job search website in India, for AI workers, with our partners developing different AI workers in the future. We recently raised a $4 million round from investors who are industry veterans. With this round of funding, we are focused on building the right team. We are pleased with how the team is shaping up and are confident in our ability to take our company to the next level,” Animesh said.

Next move

Sharing their plans, Animesh and Sanjeev said, “Enterprises face two significant challenges, namely, low-profit margins and unproductive workforce; and they will undoubtedly face tough times ahead. Fortunately, the industry is aware of solutions to these challenges. Our vision for the future involves inviting partners to our marketplace, where they can access recruiters, AI analysts, and other AI workers to optimise the performance of their employees. The platform will be available for any enterprise, large or small.”

“We aim to bring the Silicon Valley to Pune or India. We hope to revolutionise the industry and empower enterprises to thrive despite the current economic climate,” the duo said.