AI infrastructure startup Tensormesh has raised $4.5 million in seed funding led by Laude Ventures, with participation from database pioneer Michael Franklin. The company is emerging from stealth to commercialize LMCache, an open source tool known for reducing AI inference costs by up to tenfold.
Tensormesh’s core innovation lies in its handling of the key-value (KV) cache, a memory system often discarded after each query. By retaining and efficiently reusing this data, the company enables models to process information faster and more cost-effectively without increasing GPU demand.
The approach is especially valuable for chatbots and agentic AI systems that rely on continuous context retention. With its academic foundation and strong technical team, Tensormesh aims to become a key enabler of high-efficiency AI infrastructure.
