Small language models for voice and multimodal AI at enterprise scale
Smallest.ai builds compact language models optimized for low-latency voice, text-to-speech, and speech-to-text workloads in regulated environments. The stack—Python, Kafka, WebRTC, Twilio, Kubernetes, and Terraform—reflects a production-heavy, infrastructure-first engineering org. Active projects span training data engines, voice agent development, and telephony integrations, while pain points cluster around data quality, multilingual handling, and reducing iteration cycles—all critical for shipping production speech models at scale.
Smallest.ai is an AI research lab focused on compact, high-accuracy models for voice and multimodal applications across enterprise sectors. The platform operates with on-premises and private-cloud deployment options, meeting SOC2, GDPR, HIPAA, and PCI compliance standards for regulated industries. The company is headquartered in San Francisco with 51–200 employees and was founded in 2023. Revenue generation and GTM infrastructure remain early-stage priorities, reflected in concurrent hiring across engineering, data, sales, and marketing.
Python, Kafka, Kubernetes, AWS (EKS, ALB, NLB, VPC), Docker, Terraform, Helm, Argo CD, and WebRTC. For voice: Twilio and WebSockets. Observability via New Relic and CloudWatch.
Real-time data pipelines, training data engines, voice agent systems, telephony provider integrations, and GTM infrastructure. Focus areas include data quality, multilingual model support, and low-latency inference.
Other companies in the same industry, closest in size
smallest.ai's technology stack, projects, and hiring signals are inferred from public hiring and company data — career pages, public listings, and company web presence — then clustered and de-duplicated. Figures are estimates that refresh over time. Read our full methodology →
This is not an official vendor or customer list. It is a technology-adoption signal inferred from public data, intended for B2B research.