Advantageous to build new AI applications on top of LLMs: Accel's Kirani

In 2023, the world's early AI applications and large language models (LLMs) emerged primarily from the Bay Area, driven by highly specialised PhDs in deep learning. India then lagged in terms of access to these expert resources. A year later, the India story has changed dramatically.

The year 2024 will go down as a turning point with engineers in India beginning to develop innovative AI applications across various verticals, without needing deep learning PhDs. The number of AI startups building solutions on top of LLMs has since surged significantly.

India is rapidly becoming a global hub for AI applications, says Shekhar Kirani, a partner at global venture capital firm Accel.

 

He points to specific examples that demonstrate India's unique advantage in AI. A Bengaluru-based engineering team, for instance, built an AI application that automates 10-Q filings for US companies. Traditionally handled by accounting firms, the process took several days. However, the team trained a deep-learning model on top of open-source LLMs, reducing the time needed to generate these filings to under an hour.

According to Kirani, India's vast services sector encompassing call centres, accounting, healthcare management, IT, and document processing offers access to a rich dataset for training AI models. India is a goldmine for trainable data, and our engineers are perfectly positioned to build industry-specific AI applications.

LEVERAGING PROPRIETARY DATA FOR AI INNOVATION

A standout example of India's growing expertise in AI is CareStack, a dental practice management software company in Accel's portfolio.

CareStack's engineers in Kerala developed an AI application that records voice annotations from dentists and automates data entry into patient management systems. This solution, built on OpenAI's LLM, was made possible because the company had amassed a large dataset of voice samples from its dentist customers.

Similarly, Zenoti, another Accel-backed company, developed an AI-powered front-office management system for spas and salons. The AI agent handles appointment scheduling, cancellations, and customer queries in real-time. These examples, says Kirani, illustrate how Indian companies are combining proprietary datasets with existing LLMs to create solutions that are transforming industries.

INDIA'S EDGE IN BUILDING AI APPLICATIONS ON LLMS

In Kirani's view, India's combination of engineering talent, industry expertise, and access to trainable data gives the country a clear advantage in building AI applications on top of LLMs.

We are seeing small, young teamsengineers just three years out of collegecreating phenomenal AI applications, he explains.

Kirani predicts that India will lead the way in developing middleware and application layers for LLM engines, integrating and personalising AI to meet enterprise needs.

However, he also acknowledges that India faces significant competition in building the core LLM engines. The West is investing billions of dollars in this space. It's unlikely that India will outpace these global players in LLM engine development in the short term.

Instead, India's strength lies in its ability to adapt and integrate these engines, making them work for specific enterprises and personalising AI applications for users.

AI'S IMPACT ON INDIA'S SAAS AND IT INDUSTRIES

According to Kirani, the fusion of AI with India's services and SaaS industries presents a massive opportunity. India's $ 250 billion IT services industry is already partnering with SaaS startups to offer AI-powered solutions on an outcome-based pricing model. AI is not just revolutionising SaaS startups but is also reshaping the rules for the services industry.

Much like the disruption we saw in global telecom prices that fuelled the growth of India's services companies, we are now witnessing AI's ability to elevate Indian SaaS companies to the next level, Kirani explains.

He believes that India's engineering talent, combined with the right business model innovation, could make the country a shining global leader in AI productized services over the next decade.

A NEW WAVE OF TECH DISRUPTION

Kirani predicts that the next wave of tech disruption is just beginning. Companies that fail to adapt to this new reality will perish, while a new generation of startups will emerge to ride the AI wave.

At Accel, we are betting big on this wave, investing heavily in AI-based productized services. The future is bright for Indian startups that embrace the fusion of AI and services, Kirani says.