Tensormesh

Tensormesh builds enterprise-grade caching infrastructure for large language models. Its stated aim is to reduce AI inference costs and latency by up to 10x, addressing two of the more persistent constraints on running LLMs in production. The company operates in the AI/ML infrastructure and enterprise software sectors, serving teams deploying LLM-based systems at scale.

The core product is caching infrastructure designed for large language models. According to the company, it can deliver up to a tenfold reduction in both inference costs and inference latency, with an emphasis on reliability appropriate to enterprise deployment. The technical work spans AI inference optimisation, LLM caching, ML infrastructure and performance engineering.

The company describes its focus as building high-impact infrastructure for major pain points in production LLM systems - cost and latency - and as engineering for performance, reliability and the demands of real-world deployment. Further details, including founding year, team size and funding, are not publicly stated.