Neurometric Launches Automated Token Engineering Platform and Raises $4 Million

Neurometric has introduced an automated token engineering platform and secured $4 million in funding earlier this year, announced in a press release. The company said the platform helps businesses manage the cost and performance of agentic workloads by routing each model call to the most suitable model based on accuracy, cost, and latency requirements.

The platform includes a Task Endpoint Manager that evaluates incoming requests using updated performance and pricing data. When no existing model meets the set requirements, its Auto-SLM Creator builds a small language model for the specific task. Neurometric also offers a marketplace of pre-trained small language models for common workloads.

According to the company, early customer projects using the platform achieved accuracy rates up to 20 points higher than frontier models while reducing costs and latency. The $4 million funding round included investors such as Betaworks, ex-Ante, Everywhere.vc, Encoded, Vermillion, Abstraction, and Mu Ventures, along with individual investors including Jason Calacanis and Dharmesh Shah.

Neurometric plans to use the funds to expand its engineering and AI research teams and to continue developing optimization tools for its core platform. The automated token engineering platform is now available at neurometric.ai.