Aye AI, Professor
Did you know that Malar, the viral autonomous AI university professor, has a connection with the 2015 Malayalam hit film Premam? Malar—capable of teaching college students through WhatsApp prompts in the local language—has been named after the guest lecturer-character played by Sai Pallavi in the film. The pop culture connection ends there. The brainchild of Arjun Reddy, Deepika Loganathan and Aravinth Ramesh, co-founders of HaiVE, a Singapore-based startup with an office in Chennai, the artificial academic was invented to find solutions without overwhelming users with AI experience.
The result of an advanced AI-driven project aimed at revolutionising the way individuals interact with technology for educational purposes, Malar utilises state-of-the-art machine-learning algorithms, natural language processing and cloud computing technologies. “We use a customised programme to assimilate specialist knowledge base necessary to learn and teach Engineering, Management and Law to our users,” says Reddy. Now, a Tier-III student can use his or her WhatsApp—Malar’s primary mode of teaching—to get the best explanation to an academic problem that was earlier only available in Tier-I institutions.
The team wanted to bring the power of AI in the form of a solution to people in their local language and also keep it accessible. “Someone may have an iPhone 15 Pro or the latest Samsung Galaxy Flagship phone, or even a Rs 3,000 5G phone, but what they all have in common is WhatsApp. So, we launched a WhatsApp AI service,” says Loganathan. The first 20 questions per day are complimentary, after which premium services must be paid for.
Trained entirely on the engineering syllabus from Anna University, Malar can explain complex concepts in all Indian local languages. In June, the startup struck an official partnership with the university and is also open to partnering with other varsities to personalise the AI professor. The startup is in talks with the Mauritius government on how to help their student population learn through AI. The team had no idea that Malar would go viral. “We were scrambling to amp up resources when we found we got 600k-plus unplanned users. We have settled at 100k-plus daily active users now, and are growing at a good rate since we add features and knowledge bases,” says Ramesh.
However, as with all new projects, Team Malar too faced significant challenges in ensuring the accuracy and reliability of the responses of the prof. They had to fine-tune algorithms and conduct extensive testing to be sure that AI could provide precise and helpful answers.“We run an AI incubator called AI Venture Factory at St. Joseph’s College of Engineering and Technology in Chennai. The management was gracious enough to recruit interns to help us refine Malar.
Another challenge was to integrate Malar seamlessly with various educational platforms and see to it that the programme could handle a large volume of user interactions without compromising on performance. A dilemma was inventing most of the tech ourselves so that we could offer a state of the art experience while keeping in mind the computing costs,” says Reddy.
In a few years from now, one-on-one personalised coaching could very well be the norm. “We want to level the playing field for all Indians and catch up with the national vision,” says Ramesh. Will it end in a Star Trek exploration of the universe or will it get dystopian, closer to Mad Max? How the new generation uses AI learning and where our society will go from there, is a Matrix of possibilities.\