Nvidia invests $14 million in Indian AI inference startup Simplismart, valuing it at $100 million
The Series B funding backs software that lets companies run models on fewer GPUs, promising cost and performance gains for AI workloads.
Nvidia has led a $14 million Series B round in Simplismart, an Indian startup that builds software to make AI inference more efficient. The financing, which also includes Accel, Shastra VC, Micromax and Dallas Venture Capital, puts the four-year-old company at an estimated $100 million valuation.
Simplismart’s platform sits between machine-learning models and the underlying compute resources, automatically configuring deployments to meet specific speed, cost and capacity targets. By adjusting how models are served, the software can reduce the number of GPUs required for a given workload, allowing customers to extract more value from existing hardware. The solution can be run on Simplismart’s own cloud, on public clouds, or within on-premises data centers, and it offers tools for model fine-tuning, dynamic scaling and performance monitoring.
The startup cites a 56 percent reduction in GPU serving costs for AI video-generation platform InVideo as a flagship result. Other clients include video-AI firms HeyGen and Higgsfield, enterprise AI startup Ema, and Indian cloud provider Yotta, which uses Simplismart’s technology to power a serverless AI inference service in its Shakti Studio offering.
Revenue growth has been rapid. Simplismart’s annualized run-rate has risen 14-fold over the past six months to an eight-figure figure, based on actual customer usage rather than contracted recurring revenue and excluding pass-through infrastructure fees. More than half of that revenue now comes from the United States, even though the company’s customer base is larger in India. The firm employs 58 people and was founded in 2022 by former Oracle engineer Amritanshu Jain and former Google engineer Devansh Ghatak, who met while studying at BITS Pilani.
Beyond capital, Nvidia is collaborating with Simplismart on technical integrations, notably around Nvidia’s Dynamo inference stack. Engineers from both companies are working together, and Nvidia has leveraged its ecosystem to introduce Simplismart to potential buyers and partners. While the partnership may appear contradictory, Nvidia’s hardware is being used more efficiently with fewer chips, Jain argues that lowering total cost of ownership makes Nvidia’s GPUs more attractive and can accelerate broader AI adoption.
Simplismart operates in a crowded AI inference market that includes Nvidia’s own software suite and the inference services of Amazon Web Services, Google Cloud and Microsoft Azure, as well as a wave of specialist startups. The company differentiates itself by tailoring performance and cost trade-offs to each client’s service-level agreements, rather than offering fixed “fast” or “cheap” presets.
The fresh capital will fund a push into the U.S. market, additional hiring and further development of the inference platform. The announcement follows earlier reports that Simplismart was courting a $40-$50 million raise at a $300-$400 million valuation; Jain declined to comment, noting the Series B round is closed and not being extended. The next steps will test whether the company can scale its efficiency-focused model in a segment where hardware cost pressures remain a key driver for enterprise AI deployments.