Nvidia Backs Indian AI Inference Startup Simplismart in $14M Series B

Nvidia has invested in Bengaluru-based Simplismart, a startup whose software helps companies run AI models more efficiently and sometimes on fewer GPUs. Dallas Venture Capital led the $14 million all-primary Series B, joined by Nvidia, Accel, Shastra VC and Micromax, valuing the four-year-old company at about $100 million.

Simplismart’s software sits between AI models and computing infrastructure, automatically tuning response speed, request capacity and compute consumption to match each customer’s needs. It can run on its own platform or within existing cloud and on-premises setups, and offers fine-tuning, autoscaling and monitoring tools. Customers include InVideo, which cut GPU serving costs by 56%, HeyGen, Higgsfield and Ema, while Indian cloud provider Yotta uses its technology for serverless inference.

Co-founder and CEO Amritanshu Jain said companies need inference at speeds and costs suited to their applications rather than the fastest or cheapest option. He said Simplismart wins where performance and cost requirements are strict, and argued that more efficient inference ultimately benefits Nvidia by driving AI adoption. Nvidia is also collaborating on integrations, including its Dynamo software.

Annualized revenue run rate has grown 14-fold in six months to eight figures, with over half from the U.S. Founded in 2022 by Jain and former Google engineer Devansh Ghatak, Simplismart will use the funds to expand in the U.S. and develop its software.

Featured image: Credit: Simplismart