IBM Partners with Deepgram to Bring Real-Time Voice to watsonx Orchestrate

IBM and Deepgram have announced a strategic partnership to integrate Deepgram's speech technology into IBM's watsonx Orchestrate platform, designating Deepgram as IBM's inaugural voice technology partner.

Under the terms of this agreement, Deepgram's speech-to-text and text-to-speech capabilities will be embedded within IBM's platform designed for constructing and managing AI agents and automated workflows. This integration is intended to equip digital agents with enterprise-grade transcription, real-time captioning, and natural voice interaction functionalities.

The companies state that this collaboration enables enterprises to build voice-enabled agents and workflows on a real-time, scalable basis. Use cases span customer support, call analytics, and voice-driven data entry.

This partnership underscores the rising demand for conversational interfaces in enterprise environments as organizations aim to streamline interactions with software agents. Grand View Research forecasts the global speech and voice recognition market will expand from $20.2 billion in 2023 to $53.7 billion by 2030, growing at an annual rate of nearly 15%.

Deepgram's technology addresses challenges such as background noise, diverse accents, and authentic conversations. It supports 35 languages, including various Arabic and Indian dialects, and offers custom tuning and real-time captioning options. The company claims to deliver over 90% accuracy in production with latency under 300 milliseconds.

Deepgram provides its speech-to-text, text-to-speech, and voice-to-voice services via API. Over 200,000 developers utilize its cloud and self-hosted APIs. The company reports having processed more than 50,000 hours of audio and transcribed over 1 trillion words for clients including independent software vendors, enterprises, and partners.

This integration enhances the capabilities of watsonx, a suite of AI products comprising tools for building, training, and managing foundation models, as well as orchestrating AI agents in business processes. IBM positions watsonx as a core strategic component, aiming to deliver open, enterprise-focused AI integrated with hybrid cloud environments.