Founded in 2017 by Ganesh Gopalan and Ananth Nagaraj, is a Bengaluru-based startup that claims to facilitate over 1 million daily interactions with its product line, aimed at contact centers. is for

The company has more than 100 customers in the banking and financial services, insurance, telecom, automotive and healthcare industries.

In May, the startup launched a series of voice-first SLMs (small language models), carefully trained on millions of audio hours of proprietary audio datasets and billions of Indian language conversations.

“What we’re doing is something different, like combining voice and text models. It’s a multimodal model but right now we’re focusing on voice and text,” CEO Ganesh Gopalan said. told the aim In an exclusive interview

So far, the company has built a series of five models designed for the banking, finance, security and insurance (BFSI) sector.

Gopalan reveals that there are models. Multilingual. In the US, it supports both English and Spanish, while back home, the model is designed to support 12 Indian languages.

“We also plan to launch a model designed for the automotive industry as we have a lot of customers in that industry. Healthcare is going to be a sector in the future,” said Gopalan.

Building for the edge

The size of the SLMs produced by is relatively small compared to LLMs such as GPT-4 or even smaller LLMs such as Llama-3 with 7 billion parameters. Gopalan believes that the size of SLMs will come down even further, allowing them to be deployed at the edge.

“In the future, we will deploy these models at the edge as the size also comes down significantly. We believe that solutions to many enterprise problems are not always found in the typical 100+ parameter models that companies often tout. .

“These models are great for general applications but may not always effectively meet specific enterprise needs,” he said.

Additionally, many institutions are reluctant to adopt third-party models due to concerns about their proprietary nature, uncertainty about the training data used, and security concerns. Running a model at the edge where the customer data resides solves many of these problems.

“We think the models that run on the edge will be very cheap. So, at some point, all these models will be on the edge. And that’s something we’re actively working towards,” Gopalan said.

What considers its strength is its ability to quickly fine-tune the model based on enterprise data and prepare it for production.

“Having an SLM for the BFSI sector is one thing, but the real value for a company is when you have a model built just for their data. So that’s what we do. We take our model to enterprises. go and help them build on top of it. We provide them with the necessary tools to quickly launch models based on their data,” said Gopalan.

AI agents are coming.

Today, AI agents are understood The next big iteration in AI Earlier in the cycle, Kailash Nadh, CTO of Zerodha, said that AI agents are very likely to happen but maybe not in a good way.

While Gopalan agrees. AI agents The next big leap will be, he said, adding that Ganani has been building AI agents for more than four years.

“We automate processes that are performed by contact center agents through our voice bots. We built AI agents that helped one of India’s largest banks over a billion dollars in the last six months. helped with the collection,” said Gopalan.

The bot handled payment reminder calls to customers, ensured timely payment and helped them find suitable payment methods.

“For the US client, our AI agents are supporting contact center agents by providing immediate responses to customer queries as soon as the conversation begins,” he added.

AI agents will also change the way the contact center works. According to him, the future is multimodal. For example, if you have a problem with your laptop today, calling the contact center requires providing a tag number and describing the problem, which can be difficult.

“Instead, you can show a video, and an AI bot can analyze the visuals, identify problem areas, suggest solutions, and present options to the agent. The agent, human Using intelligence, it can then assess the problem and provide a solution,” he said.

IVR, the biggest bottleneck in customer experience

However, there are many concerns about creating AI agents. Contact center agents redundant. Gopalan believes there is a long way to go before AI makes call center agents redundant, if ever.

“We currently deploy AI bots for a number of use cases. However, systems are still evolving for scenarios that require human intelligence or empathy. In these areas, AI bots are yet to be developed. There are not,” said Gopalan.

Additionally, they believe that interactive voice response (IVR) may be the biggest barrier to customer experience.

“No one enjoys navigating through automated menus and pressing buttons when contacting a contact center with an urgent issue,” he pointed out.

Gopalan believes that AI will have to replace IVR before it can eliminate the jobs of call center agents. The contact center business is one segment being affected by creative AI, and talk of making AI redundant for contact center agents is widespread.

“Another reason contact centers will remain relevant for a long time is that it’s not just about getting people to answer calls. Many companies lack integration with CRM systems, ticketing tools and other necessary infrastructure. Additionally, “A significant portion of customer service knowledge resides in the minds of contact center employees,” he concluded.

Source link