One such firm is Noida-based AI startup CogniTensor , which helps companies with predictive decision-making and analysing quantifiable business solutions.
Its co-founder Ashish Airon says he got the idea to set up this company when he was doing his master’s in computer science in the UK. “What you see today was my thesis during my master’s in computer science at Oxford in 2018. The entire idea was transformed into a company. We haven’t stopped yet, and the idea is still evolving,” says Airon, who has specialised in deep learning.
However, Airon’s entrepreneurial journey in AI didn’t start with CogniTensor, but MetaTensor in London in 2016-17. The enterprise was created with the idea to connect researchers globally to build AI models that can use vast amounts of data efficiently. The initial funding for MetaTensor was provided by Airon’s father.
Two years later, in January 2018, Airon returned to India and co-founded CogniTensor with industry veterans and now fellow company directors Arun Aggarwal (59) and Pankaj Mathur (59). While Aggarwal is currently the company’s CEO, Mathur is the President. Airon heads the product & strategy division.
Sharing his experience, Mathur, who has over three decades of professional experience with big names in technology such as IBM and HP, says he saw a big opportunity in the integrated AI platform that Airon had created. “It was then much ahead of its time,” Mathur says.At that time, Airon was already being mentored by Mathur’s long-time friend Aggarwal who comes with an extensive experience in international sales, business development, and strategic partnerships at companies such as Comptel India and IBM. “We came together to create CogniTensor since Arun and I understood the market better, while Airon had a grasp over technology,” says Mathur.On a lighter note, Mathur adds: “Even the combined average age of the co-founders of the organisation was 45 — which was exactly the same that a Harvard research had suggested for a startup to be successful. So, young Airon also helped us to bring down the overall combined age of the co-founders.”
The company started its operations from a small space in East Delhi and was incorporated as CogniTensor Technology Private Ltd on 11 April 2018.
CogniTensor’s 3 ‘D’ solutions
CogniTensor uses DeepOptics, which is a low-code AI platform, to power digital transformation for enterprises by providing an end-to-end horizontal environment for faster development and deployment of automated workflows, analytics and AI or machine learning (ML) applications. “This enables us to build and deploy fully customisable and end-to-end cross-industry solutions to tackle various key business tasks,” CogniTensor’s website says.
The company provides solutions to industries such as education, healthcare, sustainability, manufacturing & distribution.
“We provide 3 D solutions–Digitisation, decentralisation and decarbonisation,” says Airon, 30.
He explains that the company provides solutions to many retailers and energy firms to help procure raw materials at the right prices. Their technology allows these companies to transform data insights into information.
“The comes decentralisation, which is about making decisions on procurement, payment and manufacturing after the data is sorted. It is the AI segment of this entire process. The third is decarbonisation, which focuses on making businesses sustainable by reducing CO2. All this is delivered through DeepOptics.”
‘Can’t compromise on pricing’
“We work on a model which is far more developed and established compared to others who are still fairly young in the ecosystem,” says Airon, who has been featured on the list — Forbes 30 Under 30 Asia in Enterprise Technology in 2020.
But when it comes to product pricing, the lanky entrepreneur admits that they are slightly on the higher side. “We can’t survive with lower pricing. We are very clear in our minds that bringing down the prices will have a negative impact on our R&D, and the product’s quality. In that case, we will not be able to serve our customers better. We cannot operate successfully below a certain threshold,” adds Airon.
Airon says that their DeepOptics tool supports a wide range of algorithms for data analysis and modelling including popular deep learning algorithms such as convolutional neural networks (CNNs) for image recognition, recurrent neural networks (RNNs) for sequence data analysis, and generative adversarial networks (GANs) for data generation. Furthermore, it supports classical machine learning algorithms like decision trees, support vector machines (SVMs), and random forests.
The company refused to disclose its revenue and profit numbers.
According to the financial data available on the business intelligence platform Tofler, CogniTensor generated a total revenue of Rs 109.16 lakh in 2019-20; Rs 45.21 lakh in 2020-21 and Rs 68.80 lakh in 2021-22. The AI company posted a net loss of Rs 16.19 lakh and Rs 16.51 lakh in the financial years 2019-20 and 2020-21, respectively, and net profit of Rs 0.18 lakh in 2021-22.
“The business scenario was challenging during the COVID-hit years. But now things are better. We aim to cross $1 million revenue by the end of 2023,” says Airon.
The company’s R&D centre is in Noida. CogniTensor has served more than 20 clients in India, Europe, and the US and has just started its Middle East operations. It has over 50 employees in India.
ET Rise sought feedback on CogniTensor’s AI solutions from its clients, but they refused to comment. CogniTensor has two strategic investors, Shell and CIIE. It is part of the Accenture Sustainability Value Accelerator at CIIE, which aims to accelerate startups with technology solutions.
“We raised around $250,000 from both Shell and CIIE combined in 2022,” says Airon. On the company’s current valuation, Airon says, “We have not disclosed it yet.”
‘Data security should be the top priority’
While experts are upbeat about the increasing demand and desire for AI solutions among enterprises, there is a concern about technology challenges such as the need for high quality training datasets and sufficient infrastructure for training and deploying AI solutions.
“For Indian enterprises, AI adoption is as much a function of the business environment as it is about the technology,” says Anusha Rammohan, Co-Chair, AI Working Group, IET Future Tech Panel.
Rammohan stresses on the importance of organisations having a strong culture of research and technology innovation with a focus on outcomes rather than tools. “Realising our true potential in AI adoption will require a robust innovation ecosystem underpinned by effective partnerships between academia, start-ups and large and small enterprises,” she adds.
A few other AI experts, ET Rise spoke to expect upcoming AI firms to highly prioritise and address data security concerns when dealing with sensitive customer information.
Airon says that data security is a top priority for CogniTensor. “We employ robust security measures to protect sensitive customer information. This includes implementing encryption techniques to secure data transmission and storage, strict access controls and authentication mechanisms, regular security audits, and compliance with industry standards and regulations such as GDPR,” says Airon.
