Nvidia in Talks to Lead $20 Million Round for Simplismart, Valuing AI Inference Startup at $100 Million

Nvidia is in advanced talks to lead a $20 million Series A for Simplismart, an AI inference optimization startup based in Bengaluru and San Francisco, at a valuation of approximately $100 million. The projected valuation would be a fourfold jump from the roughly $25 million at which Simplismart raised $7 million in October 2024, according to Economic Times, citing people familiar with the matter.

Existing investor Accel is expected to participate, with at least one new investor likely to join. Neither Nvidia nor Simplismart responded to Economic Times’ queries.

What Simplismart Does

Founded in 2022 by former Oracle and Google engineers Amritanshu Jain and Devansh Ghatak, Simplismart builds software tools that help enterprises deploy, manage, and optimize AI models in production environments. Its inference-focused platform is designed to improve GPU utilization and reduce the cost of running generative AI applications at scale, according to Economic Times.

The platform supports large language models, small language models, vision-language models, speech recognition systems, and text-to-image and video models. Customers include Tata 1mg, Mindtickle, InVideo, and Dashtoon.

Earlier this year, Simplismart announced its inference platform would be available on Nvidia infrastructure and has been working with Nvidia Inference Microservices (NIM), which allows enterprises to deploy containerized AI models as managed production endpoints with governance and cost controls.

Nvidia’s India AI Infrastructure Push

The investment fits into Nvidia’s broader expansion in India’s AI ecosystem through cloud partnerships, startup collaborations, and direct investments in enterprise AI infrastructure companies. In FY26, approximately 22% to 24% of the 1,020 total deals tracked by Tracxn were in AI/ML startups, per Economic Times.

The deal adds to a growing pipeline of large AI-native funding rounds in India. Sarvam AI is reportedly in talks to raise $300 to $350 million. Emergent recently raised $70 million led by SoftBank and Khosla Ventures.

Why Inference Optimization Matters for Agents

The timing of this investment aligns with a clear market signal: as autonomous agent workloads scale, inference cost management is becoming a critical infrastructure layer. Salesforce’s Marc Benioff projected this week that his company will spend $300 million on Anthropic tokens in 2026, and called for an “intermediary layer” to route tasks between expensive frontier models and cheaper alternatives. That routing and optimization is precisely the category Simplismart occupies.

For enterprises running fleets of AI agents, the economics of inference determine whether agent deployment is viable at scale. Simplismart’s pitch, improving GPU utilization and cutting per-inference costs, directly addresses the cost barrier that separates pilot deployments from production rollouts.

Other investors in Simplismart include Snapdeal founder Kunal Bahl’s Titan Capital, Dallas Venture Capital, LetsVenture, and Notion co-founder Akshay Kothari.