How Gnani.ai Is Enabling AI-Based Customer Support With Its Proprietary LLM Stack

A recent research report projects the global chatbot market to grow at a CAGR of 23%, reaching $42.83 Bn in the next 10 years. Now, with the advent of GenAI, this market is expected to accelerate at a faster pace, with audio and video bots poised to witness massive adoption across various sectors.

In India, too, numerous businesses across industries are leveraging AI chatbots to enhance customer support efficiency, detect fraud, and optimise costs, with no end in sight to the AI adoption mania.

To cater to the rising demand for AI chatbots, Bengaluru-based Gnani.ai has developed a conversational AI suite of products and solutions not only for specific domains but also for varied use cases. 

Founded in 2017 by Ganesh Gopalan and Ananth Nagaraj, Gnani.ai’s AI customer experience platform helps automate conversations with its ‘Automate365’ virtual assistant. 

The startup also offers Assist365, a product that can guide agents in real time as they speak with customers. It has implemented GenAI in both text and voice formats, and it offers voice biometrics for fraud detection called Armour365. 

Gnani.ai either customises its products as per each domain requirements or allows its customers to customise the products themselves as per their specific use case.

Operating in a highly competitive space, Gnani.ai distinguishes itself by focussing on niche business categories and addressing challenges in the GenAI space, such as hallucinations and inaccuracies.

The Genesis Of Gnani.ai

With more than two decades of experience in the deeptech industry and having worked with the likes of Texas Instruments and IBM, Gopalan had foreseen the impending opportunity in the AI chatbot market when several of his colleagues were getting poached by Amazon Alexa.

Soon, Gopalan and his former colleague at Texas Instruments, Nagaraj, decided to build an Indian product-led deeptech startup, paving the way for Gnani.ai. 

The startup claims to be the first venture to have developed speech engines for Indian languages.

Gnani.ai deployed its first product — an AI-based voice bot for the lending industry — in 2018. The product aimed at resolving the non-performing asset problem for a major bank.

In 2019, Samsung Venture Investment Corporation, the investment arm of Samsung, invested in Gnani.ai, along with three other Indian startups. 

The $1.68 Mn fund infusion opened a new door for the Bengaluru-based startup, as the international electronics giant had its eyes on using Gnani’s ASR (automatic speech recognition) engine to power its virtual assistant, Bixby. 

Since then, Gnani.ai has made significant strides in building custom-tuned large language models (LLMs) for domains like BFSI, insurance, healthcare, and FMCG. It also has its hands in ultra-niche categories like accounts receivable loans. Notably, the startup got a big boost after it actively adopted GenAI in 2022. 

Today, Ganani.ai claims to be one of the few players that have successfully integrated GenAI in voice bots.

“While the startup has an omnichannel application, where its AI-based virtual assistance can automate conversations across all digital and conventional communication channels, including telephony, email and WhatsApp, around 90% of its business comes from voice,” Gopalan said.

Gnani.ai today has speech engines for 10 Indian languages and it works with approximately 40 lending companies in India, including the Baja Group, TVS Credit, Muthoot Finance, and Fibe (formerly Early Salary).

Overall, the startup has a customer base of over 100 companies, which also includes a few large enterprises in the US. It has more than 12 patents in its kitty.

How Does Gnani.ai Leverage GenAI?

Speaking about the impact of GenAI, Gopalan explained how the tech enables Gnani.ai to address complex user queries effectively. Unlike traditional intent-based bots, GenAI enables the identification and resolution of specific queries.

“Before GenAI happened, let’s suppose you wanted to buy a shoe and asked the chatbot if it could get you a shoe that’s suited for hiking… an intent-based bot would find the use case and associate that with the shoes that are available in the store. Hence, we would present the right shoe to the customer,” Gopalan said.

However, Gopalan added that if the customer were to come with a request to find a shoe for an arthritis patient, the bot would fail unless the data regarding the features of an arthritis patient were coded in the natural language understanding (NLU). GenAI has been able to solve such problems.

Gnani.ai started playing with the GenAI models much before the launch of ChatGPT, Gopalan claimed, adding that the startup uses it more like an orchestration.

“We solve user problems using what we call an orchestration. Our system can decide whether it should be the intent-based bot or the GenAI LLMs that can answer specific questions, based on the user needs, while also making sure that the data is accurate and doesn’t hallucinate,” he said. 

Moving on, accuracy and data protection are other key aspects that Gnani.ai heavily bets on. 

As per the founder, accuracy and data protection are Gnani’s top priorities, which have also helped the startup foray into the highly competitive and legally sensitive US healthcare industry.

The company’s voice biometrics product, which is available with its other products as an optional product, further ensures the detection of any fraud, which is particularly beneficial for the BFSI sector. As a SaaS startup, Gnani.ai charges based on customer usage, which is the number of conversations via its platform.

Without disclosing its earnings, Gopalan said that the company has grown 20X over the last three years, and it is all set to report a profitable year in FY24. It claims to facilitate over 1 Mn conversations daily.

What’s On The Horizon

Having already expanded globally, Gnani.ai is now aiming to penetrate deeper into the US market given the startup can earn more from this market compared to India.

With GenAI adoption ramping up, the startup believes that it will be able to ramp up its growth further within this ever-burgeoning tech domain.

Besides Samsung Ventures, Gnani.ai is backed by angels, including Lakshmi Narayan, and BVR Mohan Reddy, just to name a few. Having raised around $4 Mn so far in multiple funding rounds, the startup is also aiming to raise another round of funds soon.

In India, Gnani.ai competes with the likes of Haptik and Verloop.io in the GenAI, which, as per Inc42, is projected to surpass $17 Bn by 2030, growing from a mere $1.1 Bn in 2023. While the entire industry is projected to grow at a CAGR of 48% during the aforementioned time duration, it would be interesting to register Gnani’s growth rate going ahead.

[Edited by Shishir Parasher]