Simplismart, a company specializing in artificial intelligence inference infrastructure, announced that it has raised $14 million in a Series B funding round. The round was led by Dallas Venture Capital, with participation from NVIDIA, Accel, Shastra VC, and Micromax.
The funds raised will be used for the further development of the Simplismart inference platform. This will enable enterprises to launch AI-based applications more efficiently, and cloud providers to transform their Graphics Processing Unit (GPU) infrastructure into managed AI services. Furthermore, the funding will support the expansion of partnerships with cloud providers and sovereign AI, as well as growth in North American and other global markets.
Simplismart creates a software layer situated between AI models and GPU infrastructure. Its modular inference engine allows companies to customize AI model operation according to their needs for speed, cost, and performance. The company's orchestration platform manages deployment and scaling in both cloud and on-premise infrastructure.
Over the last six months, the company's annual revenue growth rate has increased 14fold. Companies such as HeyGen, Higgsfield, InVideo, Shiprocket, Apna, and Ema run production workloads on the Simplismart platform, covering video, voice, multimodal, and enterprise AI applications.
Amritaanshu Jain, Co-founder and CEO of Simplismart, noted that while the industry has spent many years improving AI model capabilities, the advantage is determined by how these models are utilized: the cost per interaction, application response speed, and reliability under increasing demand. He added that the team spent four years creating the software layer between the model and the GPU, helping clients optimize workload deployment and operation.
An example of usage is the Exotel communications platform. After Exotel found that the economics of closed API models were unsuitable for its voice agents, the company fine-tuned an open model on its own data and migrated the inference to Simplismart. By sharding GPUs, Simplismart reduced the cost per minute for Exotel by 57% compared to closed API models. Additionally, it lowered costs below that of self-hosting the same open model and reduced response time from 1.1 seconds to less than 300 milliseconds.
As part of its offering to cloud providers, Simplismart collaborates with NVIDIA to simplify the deployment and operation of AI infrastructure at scale. Unlike inference providers who buy or reserve GPU capacity and resell it, Simplismart itself does not own computing power. Instead, it provides a software layer that transforms third-party GPU clusters into ready-to-use AI production infrastructure. Cloud providers can use this platform to offer managed AI services under their own brands without building their own software stack from scratch. This approach is already used by Yotta, where the Simplismart platform forms the basis of the Shakti Studio server inference layer.
Dayakar Puskoor, Managing Partner at Dallas Venture Capital, stated that inference is becoming one of the most significant drivers of change in the AI economic model. He emphasized that Simplismart is rapidly becoming a global player, helping enterprises execute AI workloads more efficiently and enabling cloud providers to convert GPU power into scalable services.
Devansh Ghatak, Co-founder and CTO of Simplismart, reported that the company's modular inference engine allows configurations to be adapted to the requirements of individual workloads, and the orchestration platform manages deployment across various infrastructure environments. Ghatak also mentioned that the new funding will support the development of a dynamic GPU scheduling layer, which can align AI workloads with available computing power across different regions and time zones, helping cloud providers better utilize their GPU infrastructure.
