
Bengaluru-based Agentic AI startup Karta AI has raised $3 million (about Rs 28 crore) in a seed round led by PeerCapital and Cedar Hill Capital, with participation from Wyser VC, CDM Capital and Seedstars. Founded in 2024 by former Razorpay executives Vijaypal Singh and Abhishek Kushwaha, the firm automates complex inbound customer calls for banks and financial institutions, targeting multi-step queries that span several enterprise systems.

Karta AI claims to handle roughly 20 lakh customer conversations a month across more than 25 languages, with 10-12 paying clients including PhonePe, Mindlabs, Bajaj and Aavas. It reports approximately 70% end-to-end resolution of borrower queries without human intervention and about 50% lower cost per query. Open Magazine notes the claims lack independent verification and asks whether the resolution rate covers all incoming queries or a selected set. Karta AI plans to use the capital to expand go-to-market efforts and product development over the next two years.
Inc42’s article is a neutral product-detail report focused on Karta AI’s funding, its technology stack and the founders’ background, repeating the startup’s claims without verification or external counterpoints. Open Magazine’s piece frames the same news as a test question: can voice AI really fix India’s bank helplines? It interrogates Karta’s figures, asking whether the reported 70% resolution rate covers all queries or a selected subset, whether a resolved query means information given or action completed, and how often customers return with the same problem. It also introduces comparative benchmarks from PolyAI, KeyPoint Credit Union and Klarna, none of which appear in Inc42’s coverage. Where Inc42 reports the startup’s claims as facts, Open Magazine labels them company claims whose definitions matter.
Coverage: 2 sources, 2 neutral
Sources (2): inc42.com (neutral report), openthemagazine.com (neutral report)
This brief was synthesised by AI from the 2 sources linked above, so one read covers every framing they carry. Methodology and corrections.