Real-time context
The relevant facts stay available as the conversation evolves.
Invest
The company is building agents that navigate real conversations and finish multi-step work. The Dynamic Language System manages context, intent, and workflow state while the call is happening.
Traditional voice agents complete 85% of tasks on clean text and 30% on real, artifacted audio. Live calls change context, leave the script, and require more than one step. That is the product requirement, not an edge case.
The relevant facts stay available as the conversation evolves.
The path changes with intent, branches, and unexpected answers.
The system is designed to keep executing when the call gets harder.
The interaction stays natural while the context gets richer.
Market
Anywhere coordination still depends on a call: care scheduling, insurance, revenue follow-up, benefits, HR, and vendor operations. DLS is an architecture for those workflows, not a single vertical script.
Go to market
More than 50 years of combined work in machine learning, voice, and enterprise AI platforms. BabbleLabs was backed by Dell Technologies Capital and Intel Capital, delivered over 10× ROI, and was acquired by Cisco in 2020.
CEO and founder. Former Cadence and Qualcomm. Co-founder of BabbleLabs through the Cisco Webex acquisition.
COO, CFO, and co-founder. Finance and operations at BabbleLabs. Background across startups and GE Capital.
CTO and co-founder. AI research at BabbleLabs, then AI for Webex. Machine learning, enterprise software, and embedded AI.
Board member and strategic advisor. Founded Tensilica and BabbleLabs. Audio, video, language, and enterprise systems.
The reliability problem is visible. VaraiE is building the architecture, the product, and the distribution to go after it.