Top-10 Artificial Intelligence Startups in India
With 1.3 billion people, India is an incredible country with things to see that defy the imagination. Dudes that stand on one foot for twelve years smoking weed, a police force that protects millions of sacred cows that roam the city streets, and a man with 39 wives, India has no shortage of interesting things to check out. In the business world, however, India is mostly known for their less-than-adequate call centers which have helped transform the country into the outsourcing capital of the world. It’s also a land of software developers, and consequently, there is a burgeoning startup culture that doesn’t receive much global recognition – kind of like India’s artificial intelligence (AI) scene.
With a solid base to start from, the Indian government recently published a national AI strategy which correctly recognizes that India is lagging behind China and the US when it comes to AI development. While the strategy is only a framework at the moment, budget allocations are expected to follow. In the meantime, we decided to take a look at the top-10 AI startups in India that have taken in the most funding to-date.
Top-10 Artificial Intelligence Startups in India
Founded in 2003, Bangalore startup Manthan has raised $98 million from the likes of Temasek Holdings and Fidelity to develop a retail analytics platform based on big data. Taking on enterprise management juggernauts like Salesforce, SAP, and IBM, the company’s offering employs AI to provide descriptive and predictive analytics for users, recommend actions, and grow customer engagement.
The business platform has 170 clients across 21 countries including grocery stores, convenience stores, and fashion companies, all looking at inventory, pricing, promotions, marketing, and customer targeting. Additionally, Manthan launched a Natural Language Processing (NLP) engine called Maya early last year which connects to Amazon’s Alexa or Apple’s Siri and acts as a business assistant, answering questions about the business like last month’s profits, sales trends, or product groups, all the while getting better over time with the help of machine learning. Geared towards the C-Suite, Maya faces some stiff competition from startups like Noodle.ai.
Founded in 2015, Bangalore startup SigTuple has raised $25 million from investors including Accel to join the many AI companies out there that are developing medical diagnostic solutions using machine learning. According to the company, 100 million people in India alone are more than 100 kilometers away from a well-resourced medical center, and more than 400 million people are in the same situation worldwide. SigTuple’s AI algorithms can analyze patient samples, send them to a pathologist for review, then send final results back to the point-of-care in five minutes. In this way, initial diagnosis is established right away, and ambulances can be dispatched to carry emergency medicine to patients without waiting for them to arrive at a proper hospital.
The platform is being built to perform basic screening and advanced diagnosis of blood, urine, and semen samples, along with x-rays, and retinal scans. The blood analysis process has already gone through three clinical validations and is being used in closed beta testing by partners of SigTuple. The company has also created an automated microscope to make up for the lack of pathologists, and is running its own diagnostics lab as well.
Founded in 2013, Mumbai startup Haptik has raised $12 million in funding to develop specialized chatbots for consumers, service companies, and enterprises. Joining the “7 AI Chatbot Startups Giving Technology a Voice”, Haptik has devised a handful of algorithms to provide an end-to-end tool that lets you “build your own chatbot”. The company not only provides specialized bots for different industry applications, but gives clients a drag-and-drop bot builder, a human-bot hybrid interface for human-to-AI transitions, and detailed conversation analytics.
The resulting chatbots are customized precisely to the business context they are meant to serve. Earlier this year, Haptik partnered with Amazon Web Services and will incorporate AWS’s advanced database framework, cloud offering, and AI tools going forward. Haptik already works with a large global clientele including Samsung, Coca Cola, and Amazon Pay, and also offers a personal assistant app for consumers. The company’s “Named Entity Recognition” technology understands text-based datasets, and is available as an open source library to help independent developer projects.
We covered CreditVidya, a credit scoring system for first-time borrowers, last year in our article on AI fintech startups that offer loans on new credit. Founded in 2012, the Mumbai company has raised $7 million in total funding so far to “empower distribution of affordable credit to 50 million underserved [clients] by 2020”. The company uses machine learning algorithms to sift through thousands of publicly available data points on potential loan applicants to arrive at a credit score in five minutes. CreditVidya is currently developing an analytics suite to track loan portfolios as a whole. The company typically works with local financial institutions helping applicants move into the 21st century deeply in debt – just like we do here in ‘Murica.
Founded in 2016, Calcutta startup Mihup Communications has raised $6.6 million from Accel Partners to develop a voice interface that can be integrated into any smart tool using NLP. Mihup wants to tackle the problem of recognition of Indian languages and different dialects. Their speech-to-text converter and NLP algorithms understand Hindi, Bengali and Indian English at the moment. Mihup started as a personal AI assistant but has since pivoted to the business-to-business segment targeting media, entertainment, home automation, service center, and connected car clients.
Founded in 2014, Bangalore startup Charmboard has raised $5 million to develop a video tagging platform. Visually similar to Pinterest, it contains mainly Indian movie scenes, TV programs, and music videos where users can tag or bookmark anything that appears in the frame. When you click on a video frame, it is saved as a “charm” and can be shared on social media platforms or saved for your reference. Entries on Charmboard will generate suggested apparel and accessories, similar to the ones worn in the video using computer vision and – you guessed it – these contain links to shops where you can buy them.
Users can create boards of videos and charms they like, in much the same way it works on Pinterest. Charmboard grew out of the Indian startup accelerator program of British retailer Target and is gaining traction among consumers. Now you can figure out which dress Priyanka Chopra wore to the Oscars and then spend approximately six years’ salary on it so you can try to look like someone who spends most her time pretending to be someone else.
Founded in 2012, Bangalore startup EdGE Networks has raised $4.6 million to develop an AI-based recruitment and HR tool, driving another nail into the coffin of corporate “soft skills”. EdGE’s AI algorithms promise to automate hiring, workforce management, and provide analytics so you can see what a great job it’s doing. According to the company, the time to shortlist candidates is reduced by 58% and the conversion rate from recommendation to hiring moves to 28%. The system also optimizes task allocation based on employee profiles, sets learning paths for necessary skills, flags flight risks, and forecasts supply and demand, among other features. In short, it does everything your average HR department does except pester you about telling off-color jokes. The platform uses NLP to sift through thousands of applications, job portal entries, and internal profiles, to compile some “big data” for the algorithms to munch on.
Founded in 2016, Chennai startup OptaCredit raised $4 million in debt financing early this year to develop their version of an AI-based data-driven lending platform. The focus is on unsecured personal loans for salaried employees, and machine learning is used to produce a risk assessment and send a decision to applicants the same day. The company is offering loans directly to consumers and through company partnership programs as well, and has a reach of more than 1,500 employees through these partnerships. The loans are underwritten by OptaCredit and the money comes partly from the company’s balance sheet and partly from its financing partner, DMI Finance. The platform charges interest rates of 8.8%-15.8% per annum and a 2.5% processing fee on top of that, which seems reasonable compared to the central bank’s current base rate of 6.25%. They are planning to expand their reach to 160,000 customers and disburse $1 billion in loans by 2020.
Founded in 2015, Bangalore startup Playment has raised $2.5 million in funding from the likes of Y Combinator to develop a platform that is used to train computer vision algorithms for third parties. AI algorithms are as good as the datasets used to train them, and Playment is offering to turn raw data into tagged data so that computer vision AI models can make sense of it.
First, they receive the raw datasets from customers, then their own algorithms browse through the data and annotate whatever they can safely recognize. The rest is tagged manually by “humans-in-the-loop” who are crowdsourced through their mobile and web apps in exchange for a small fee. Finally, Playment runs a quality assurance algorithm on the results to make sure everything is in order. The company has a handful of autonomous vehicle and robotics customers, and also lists Alibaba Group as a client.
Founded in 2015, Bangalore startup niki.ai has raised $2.4 million in funding, partly from Ratan Tata, the Indian magnate, to develop a chatbot that helps automate things like payments and online ordering. Niki uses NLP to help you recharge your phone account, pay utility bills, or book hotels. It also doubles as a search assistant, recommending the best matches for your search criteria. The next steps in its development will be voice recognition and multilingual support. (For the record, we’re not huge fans of chatbots but they seem to be all the rage these days.)
When compared to countries like China, India’s AI startup funding seems dismal, but it is also significantly above Russia’s, another world power that entered the global AI arena recently. With the second-largest population in the world, India has a humongous local market to target. International venture capitalists like Fidelity and Accel Partners are beginning to look towards India for compelling investment opportunities, and the government’s commitment to AI reinforces the need for the world to consider India as a potential player in the AI arms race. In the words of Ghandi, “first they ignore you, then they laugh at you, then they fight you, then you win.”
Despite what the pundits say, FAANG stocks (Facebook, Apple, Amazon, Netflix, Google) don't give you real exposure to AI. Read about 7 stocks that give you true pure-play exposure to AI in our guide to investing in AI healthcare companies, freely available to Nanalyze subscribers.