Harsh Vardhan Goenka, known for his achievements in competitive programming and cloud infrastructure, has become the first founding engineer at Kalpa Labs, a Y Combinator startup developing advanced generalist audio models designed to transform audio generation beyond text-to-speech.
- Goenka transitioned from Rubrik to Kalpa Labs to work on cutting-edge audio AI.
- Kalpa Labs aims to create a single model capable of diverse audio generation tasks.
- The startup culture emphasizes broad ownership and learning across fields.
What happened
Harsh Vardhan Goenka, previously instrumental in improving Rubrik’s cloud infrastructure and testing speed, joined Kalpa Labs as their founding engineer and researcher. Initially hesitant, he was convinced to take on the challenge after a rigorous late-night interview involving a complex filesystem problem reignited his passion for deep technical problems. Kalpa Labs, a Y Combinator Winter 2025 startup, is focused on developing generalist audio models that can generate diverse audio outputs such as podcasts, singing in custom voices, and mood-driven sound synthesis.
With a background rooted in competitive programming and scalable engineering, Goenka is shifting his focus toward machine learning research—a domain new to him but fully embraced within Kalpa Labs’ culture. From day one, he took on responsibilities that span not only ML infrastructure and coding but also product design, workflow optimization, and broader product ownership.
Why it matters
Kalpa Labs is addressing a significant gap in the audio AI market by aiming to create a unified generalist model akin to GPT-3 but for audio—a field that has yet to experience a transformative breakthrough. Unlike traditional text-to-speech systems, the company envisions models capable of versatile, context-aware generation that can mimic human voice intricacies and emotional tone with high fidelity.
Harsh’s journey into machine learning research without prior formal experience demonstrates Kalpa Labs’ philosophy of cultivating multidisciplinary expertise within a small, agile team. This approach encourages fresh perspectives on challenging problems and accelerates innovation—a vital asset for startups operating at the frontier of AI technology.
What to watch next
Kalpa Labs is actively expanding its team and looking for engineers who thrive in small, dynamic startup environments and are eager to push their limits. Observers should watch how the company progresses in realizing its vision of truly generalist audio models that can revolutionize content creation, personalized audio experiences, and voice synthesis.
Goenka’s evolving role, which blends engineering, research, product input, and collaborative culture-building, will be crucial to shaping Kalpa Labs’ trajectory. The startup’s ability to iterate through challenges and maintain a strong team ethos around ownership and learning will also be key indicators of its future success.