How Sarvam AI's 'core intelligence factory' helps banking sector
What's the story
Sarvam AI, a leading player in the field of applied artificial intelligence (AI), has unveiled its latest offering: the "core intelligence factory." The company's head of Applied AI, Vedant Trivedi, believes that AI is the third biggest shift in India's banking, financial services, and insurance (BFSI) sector. He emphasized that India is still in its first year of this AI revolution.
Addressing issues
Addressing the AI chaos in BFSI
Despite being in the early stages of the AI revolution, the BFSI sector is already facing a deluge of offerings such as voice bots and document AIs.
This has resulted in scattered pilots, vendor lock-in, fragmented data, and compliance paralysis.
To tackle these challenges, Sarvam has introduced its "core intelligence factory," a modular stack that separates fast-moving layers like GPU infrastructure from those where institutions need to retain control and intellectual property (IP).
Strategic autonomy
Modular stack inspired by core banking system
Trivedi said the "core intelligence factory" is inspired by the core banking system developed over the last two to three decades.
He emphasized that these systems offer strategic autonomy to plug and play different layers of the stack as per their requirements.
The bottom layer of this modular stack relates to infrastructure, which Sarvam recommends sourcing from a secure, centralized "token factory" within India.
Management
Layers of the 'core intelligence factory'
The models layer of the "core intelligence factory" prevents institutions from getting locked into a single provider.
The governance layer deals with model risk management and a centralized "kill switch" to enforce policy compliance across AI agents at runtime.
The context engine layer determines how a bank's data is pre-packaged and represented to agents for accuracy and cost-efficiency.
Finally, the orchestration layer improves performance.