Hamming.ai, an AI voice agent reliability platform, announced that it raised $3.8 million in seed funding led by Mischief, with participation from Y Combinator, AI Grant, Pioneer, Coalition Operators, Coughdrop and angels including Hiten Shah, Ran Makavy, Max Kolysh, Richard Aberman and Kulveer Taggar.
People make billions of calls daily, and as conversational AI advances, most calls will eventually become handled by AI (built by thousands of companies tackling every single vertical). But making AI voice agents reliable is hard. A small change in prompts or model providers could cause large changes in voice agent responses. Engineers currently spend hours a day testing their agents by hand, a slow, ad hoc, and often inadequate process. Even when the voice agents are launched to production, operations teams listen to thousands of calls per day to discover edge cases with their voice agents that were missed by manual testing methods. This makes these voice AI systems costly to set up and, due to deficient testing infrastructure, potential sources of liability and adverse public attention.
Hamming automates the testing, monitoring, and governance of AI voice agents. Deploying its own AI voice agents that work like real people, Hamming.ai can simultaneously place thousands of phone calls to client voice agents. And Hamming.ai also provides LLM prompt management solutions for B2B teams, automated AI voice agent red-teaming to detect vulnerabilities and call analytics solutions to track how users engage with AI voice agents in production and flag cases needing attention. Hamming.ai's approach is 20x faster and 10x cheaper than having humans test AI voice agents manually. It aims to improve these margins through additional product development and refinement.
Co-founders Sumanyu Sharma (CEO) and Marius Buleandra (CTO) bring experience building trust and safety infrastructure at Citizen, the Founders Fund-backed personal safety network, where they met in 2019. Sharma had helped quadruple Citizen's user base as its Head of Data. Before Citizen, he grew an AI-powered sales program at Tesla to 100s of millions in revenue per year as a Senior Staff Data Scientist. He researched AI-powered medical image search at the University of Waterloo, Canada. Along wtih his time at Citizen, Buleandra comes to Hamming.ai with experience in data infrastructure, AI, and complex systems engineering at Anduril, Square, and Microsoft, and was a founding engineer at Spell, a machine-learning observability and infrastructure startup acquired by Reddit.
KEY QUOTES:
"Conversational AI is advancing rapidly -- but most testing and governance tools haven't caught up to match developer needs and compliance realities. Hamming.ai will be foundational to ensuring this technology advances safely, as AI reshapes interactions between businesses and their customers."
- Lauren Farleigh, Co-Founder and GP at Mischief