Originals · Startup · Startups & Fintech
Hunar.AI raises Rs 40 crore Series A led by Blume Ventures for voice agents that hire and manage frontline workers
The Bengaluru company, which counts Swiggy, Delhivery and Flipkart among its clients, plans to build its own inference layer and small language models for workforce tasks.
Hunar.AI, a Bengaluru based conversational AI company that works on hiring and managing frontline workers, has raised Rs 40 crore in a Series A round led by Blume Ventures. Hyderabad Angels Fund, Together Fund, Seedstars, AVNM Ventures, PhysicsWallah co-founders Prateek Maheshwari and Alakh Pandey, and MediaMint co-founder Aditya Vuchi also took part, and the company says the round takes its total funding to Rs 65 crore. Regulatory filings reviewed by Entrackr show Rs 39.3 crore raised through compulsorily convertible preference shares, with Blume putting in Rs 29 crore, and Entrackr estimates the post round valuation at about Rs 180 crore.
Founded in 2022 by Krishna Khandelwal and Shantanu Bhattacharyya, the company sells two products, Hunar Voice and Hunar Connect. Both use AI agents that talk to candidates and employees over voice calls, WhatsApp and chat to source and screen applicants, schedule interviews, onboard recruits, run training and follow up on retention. The platform supports more than 20 languages, including Tamil, Telugu and Kannada. Its listed clients include Swiggy, Delhivery, Instamart, Pronto, Blue Dart, Flipkart, Aditya Birla Capital and Tata Croma.
The company says its agents handled more than 15 million conversations in September 2026 and that volumes have been doubling every quarter. Khandelwal told ETEntrepreneur that Hunar.AI crossed $1 million in annual recurring revenue a few months before the raise. These are company figures and have not been independently audited.
The more interesting part of the plan is technical. Hunar.AI currently relies on outside foundation models and text to speech providers for some capabilities, and Khandelwal said a large part of the new money will go into owning the inference layer and building small language models for workforce management, trained on its own conversation data. He argued that this would give the company cost and quality control and faster responses. It also plans to move towards pricing tied to business outcomes rather than per user charges, and lists the Middle East, Latin America and Southeast Asia as possible markets abroad.
