Skip to content
MukhyamantriMukhyamantri
BusinessKarnataka

Ex-Razorpay duo's Karta AI raises $3 million to put voice agents on bank helplines

Bengaluru-based Karta AI has raised $3 million in a seed round led by PeerCapital and Cedar Hill Capital. The startup builds voice AI agents for inbound calls at banks and insurers and handles about 2 million conversations a month.

· · 1 min read

Bengaluru: Karta AI, a voice AI platform that handles inbound customer calls for banking, financial services and insurance (BFSI) firms, has raised $3 million in a seed round led by PeerCapital and Cedar Hill Capital. Wyser VC, CDM Capital and Seedstars also participated.

Founded in 2024 by former Razorpay executives Vijaypal Singh and Abhishek Kushwaha, the startup runs AI agents across voice, chat, email and WhatsApp. It targets inbound queries that need coordination across multiple enterprise systems rather than routine FAQs. The company says it handles around 2 million customer conversations a month across more than 25 languages, and that its agent resolves roughly 70% of borrower queries end to end without human intervention, cutting the cost per query by about 50%.

The agents are used for loan disbursal updates, statements of account, foreclosure statements and the release of property documents. Cofounder and CEO Vijaypal Singh said the startup has 10-12 paying customers, including PhonePe, Mindlabs, Bajaj and Aavas. Queries involving fraud or lacking clear standard operating procedures are routed to human agents, he said. Karta AI uses models from OpenAI and Google's Gemini family alongside open-source models such as Kimi and Gemma. The fresh capital will go into go-to-market and distribution over the next two years, with a portion for product development and emerging AI capabilities such as self-learning, Singh said. Over the longer term, the company plans to build reusable, industry-specific capabilities and expand into international markets. Karta AI targets L2 and L3 queries, which require an agent to interpret multiple customer intents and pull data from different enterprise systems. High-risk matters such as fraud, and processes without clearly defined standard operating procedures, are routed to human agents. The startup operates in a segment that has drawn rising investor interest in India, with peers such as Desible.ai, Navana.ai, Rezolv and Gnani.ai also building agentic voice AI for financial services.