A Bengaluru startup that helps companies run AI models in production is closing in on a funding round that would quadruple its valuation in 18 months, with Nvidia in talks to lead it. Nvidia is in advanced talks to lead a $20 million round in Simplismart at a valuation of around $100 million, people familiar with the matter told The Economic Times on Monday. The projected valuation would mark a fourfold jump from the roughly $25 million at which the company raised $7 million in October 2024.
The deal is not closed, and terms could still change, people close to the talks cautioned. An existing investor, Accel, is expected to follow on, and at least one new institution is likely to join the round. For Nvidia, it would be the latest in a series of moves to deepen its footprint in India’s fast-growing AI infrastructure market.
Simplismart, headquartered in Bengaluru with a presence in San Francisco, builds an AI inference platform that helps enterprises deploy and optimize models in production. The platform supports large language models, small language models, vision-language models, speech recognition systems and text-to-image and video models. Its customers include Tata 1mg, Mindtickle, InVideo and Dashtoon. The startup was founded by Amritanshu Jain and a small team of engineers who came out of the Indian enterprise software scene, and it positions itself around model deployment and optimization rather than foundation-model development.
The company has been tying itself to Nvidia’s ecosystem for more than a year. Earlier this year, Simplismart said its inference platform would be made available on Nvidia infrastructure as part of an expansion of its enterprise AI offerings, and it has worked with Nvidia Inference Microservices, which lets enterprises deploy containerized AI models as managed production endpoints with greater governance and cost control. In October 2025, the startup formed a strategic partnership with Yotta, an Nvidia cloud partner in India, embedding its inference-first platform into the serverless inference layer of Yotta’s Shakti Studio to provide low-latency generative AI services for Indian enterprises and government agencies.
The company’s October 2024 round — $7 million at a roughly $25 million valuation — was led by Accel, with participation from Shastra VC, Titan Capital and angel investors including Notion co-founder Akshay Kothari. That investor base is expected to extend into the new round. If the $100 million valuation is finalized, Simplismart’s value will have quadrupled in about 18 months — a trajectory that reflects the broader heating up of India’s AI infrastructure sector, where positioning in the GPU-based services buildout often matters more than near-term revenue.
The potential deal adds to a wave of fundraising across India’s AI ecosystem. Sarvam AI is in talks to raise between $300 million and $350 million, according to The Economic Times. Emergent recently completed a $70 million round led by SoftBank and Khosla Ventures, and Neysa and others are raising or negotiating fresh capital. According to Tracxn data cited in reports, roughly 22% to 24% of the 1,020 startup deals recorded in India’s fiscal year 2026 were tied to AI and machine-learning companies.
Nvidia has been particularly active in the country. Senior executives have publicly urged India to raise its AI infrastructure investment from around $1.2 billion to a level that can compete globally, and the chipmaker has pursued cloud partnerships, startup collaborations and direct investments in companies building enterprise AI infrastructure. Executives have said India’s data-center buildout and its large engineering talent base make it one of the most important AI infrastructure markets outside the United States and China, and the company has backed local cloud providers and model builders as a way to seed demand for its hardware.
The investor math reflects where the market’s center of gravity is moving. Training has become a game for a handful of well-funded labs, but inference — the running of trained models to answer queries — is where AI touches actual customers, and it is far more distributed. Startups like Simplismart sell the plumbing for that layer: serving stacks, optimization and cost controls for enterprises that want AI in production without building the infrastructure themselves. The fourfold valuation jump in 18 months suggests the market is pricing that opportunity early, and global capital is flowing to Indian startups building enterprise AI infrastructure rather than consumer apps.
A lead position in Simplismart would put Nvidia inside a fast-moving inference platform with an enterprise customer list already in hand — a modest check that cements its position in India’s enterprise AI pipeline. For Simplismart, an Nvidia-led round would be the strongest possible endorsement of its inference-first strategy: the chipmaker does not need the capital return, it needs customers running models on Nvidia hardware, and an equity stake aligns both incentives. Whether the deal closes, and at what terms, will be the next signal of how far Nvidia is willing to go to own the AI supply chain in India.
