Upwind Leverages Nvidia AI Technology to Enhance Cloud Security

Font: Financial Modeling Prep  • Mar 23, 2026

Upwind, based in San Francisco, is a prominent player in cloud security, founded by Amiram Shachar and the team behind Spot.io. The company has raised $430 million since 2022, backed by investors like Bessemer, Salesforce Ventures, and Greylock. Upwind's latest innovation involves using Nvidia AI technology to detect malicious AI prompts with high precision, a significant advancement in the security landscape.

The collaboration with Nvidia is crucial as traditional security measures fall short against new threats like prompt injection and data exfiltration. Upwind's approach, presented at the RSAC Conference, achieves approximately 95% precision in detecting malicious Large Language Model (LLM) prompts, maintaining sub-millisecond inference times. This is vital as enterprises increasingly adopt generative AI, with Gartner predicting over 80% will use AI APIs, models, or applications this year.

Upwind's solution uses a three-stage architecture for production environments. The first stage employs a lightweight classifier with 99.88% accuracy to filter traffic. The second stage utilizes Nvidia's nv-embedcode-7b-v1 model for semantic threat detection, achieving 94.53% accuracy with rapid inference times. The final stage involves selective LLM validation using the NVIDIA Nemotron-3-Nano-30B model and NeMo Guardrails to ensure security policy alignment.

Nvidia, trading under the symbol NASDAQ:NVDA, plays a key role in Upwind's solution. The company has a price-to-earnings (P/E) ratio of 35.67, indicating investor confidence in its earnings potential. Nvidia's strong financial metrics, such as a low debt-to-equity ratio of 0.073 and a current ratio of 3.91, reflect its robust financial health, supporting its collaboration with Upwind in enhancing AI security.

By integrating LLM threat detection into its platform, Upwind transforms malicious prompts into actionable security events within the cloud ecosystem. This collaboration with Nvidia demonstrates that organizations can effectively balance innovation and security, addressing the evolving challenges in the AI landscape.