We are excited to announce that we are leading the $6.3M seed financing in Miravoice, a specialized AI voice agent platform that automates structured, voice-based data collection for research and operations teams. Miravoice’s mission is to make it effortless to gather high-quality data from spoken conversations at scale and we couldn’t be more thrilled to partner with them on this journey.
Miravoice is led by Nishant Jain (CEO), Danny Leybzon (CTO), and Shreyas Tirumala (COO & President), who collectively bring experience from VMware, Google, WhyLabs, Apple, and McKinsey. Nishant and Shreyas were classmates at Yale, where they both majored in Computer Science, and they later teamed up with Danny – Shreyas’s longtime friend from their high school debate team days – who also shares a strong technical foundation. That history and chemistry really shows up in how they work together, and it was a major reason we were excited to invest: all three have been close for years, and it’s rare to see a founding team with this level of trust, alignment, and technical depth from day one.
If an organization has ever tried to conduct thousands of surveys or interviews over the phone, they know how painful and expensive it can be. Until now, automating these structured voice conversations has been nearly impossible due to poor AI quality and technical complexity. Early voice AI solutions were unreliable, often misunderstanding respondents or “hallucinating” incorrect information, and they required significant engineering efforts to maintain. As a result, most large-scale phone surveys are still done manually by call centers, taking weeks of work and costing a fortune (5-6 figures). Organizations have been forced into a frustrating tradeoff between quality, speed, and cost, either spending huge sums on human callers or settling for smaller samples and lower data integrity.
Miravoice’s platform breaks this tradeoff by building a research-grade AI voice agent, underpinned by a proprietary intelligence layer that makes real-time decisions during a call to keep respondents engaged and ensure every question is answered without bias. This means Miravoice can smoothly handle even the most complex interviews (the founders published research papers showing that it can conduct a 125-question phone survey without a single AI hallucination) all while maintaining an excellent participant experience. Any research or operations team can design and launch a voice survey in Miravoice without writing a line of code, avoiding the heavy developer dependency that has plagued previous solutions.
The market opportunity for Miravoice is massive with more than 45K market research firms worldwide that conduct voice surveys and interviews each year. These companies spend billions on phone-based research and are hungry for a faster, more scalable, and more verifiable way to collect high-integrity data. As organizations across industries become more data-driven, the need for an automated, reliable voice data collection solution is more important than ever.
We’re excited to back this team as they pioneer a new approach to voice-based data collection. In a world where decisions are increasingly driven by data, Miravoice’s solution ensures that organizations can gather critical information at scale without compromising on quality or integrity.
All Posts