The audio and speech recognition company taps Oracle Cloud Infrastructure to power its voice AI platform and scale to support explosive growth.
“We view this relationship with OCI as long term. We’re excited about taking advantage of the GPUs and using that to train our next generation of voice AI. There's a lot that we think that OCI will provide for us in terms of future growth.”
SoundHound has developed an independent voice AI platform that allows businesses across industries to integrate conversational voice assistants into their products and services. SoundHound’s voice AI platform has the ability to understand the complexity of human speech so that conversational experiences are more natural and intuitive, allowing people to interact with machines the way they do with each other–by simply talking.
As the company continues to expand at a global scale, it is committed to offering a reliable service with zero downtime. To achieve this, it needed an infrastructure capable of scaling along with the company, and powerful enough to process complex queries.
Why SoundHound chose Oracle
As SoundHound grew, it looked to Oracle Cloud Infrastructure (OCI) cloud native services such as OCI Functions, OCI API Gateway, and Oracle Container Engine for Kubernetes (OKE) due to speed, performance, and capacity. As it focused on building the next generation of voice AI software, having a reliable and secure cloud infrastructure was top of mind for SoundHound.
By using Oracle Cloud Infrastructure, SoundHound processes more than 100 million queries each month—a number that doubled in the first six months of 2021. It achieved 1 billion queries through the course of 2021 and expects that scale to continue. OCI offers capacity in certain regions that was not available with other cloud vendors—allowing the company to expand as it brings in more users.
Even with the global chip shortage, the Oracle team made sure that the company had the necessary allocations and resources to continue growing. Its technology is geared toward conversational experiences and real-time interactions involving voice latency. The speed and throughput provided by OCI allows SoundHound to adhere to its commitment of providing the fastest and most accurate voice AI services to its customers.
“We run billions of voice AI queries on OCI, using a mix of Kubernetes infrastructure with OKE, GPUs, HPC, and other services. We've seen a 50-60% performance boost compared to our previous cloud, along with 2x cost reduction—all while doubling our usage,” states James Hom, Cofounder and Vice President of Products at SoundHound.