Add DeepKeep to Commercial MLSecOps and AI Security Tools

[gilarel]

Important

Read the contribution guidelines before submitting.
Awesome MLSecOps is a curated technical resource, not a product directory.

Submission

• Resource: DeepKeep
• • Canonical URL: https://www.deepkeep.ai
• • Suggested section: Commercial MLSecOps and AI Security Tools
• • Project launch date: Publicly available; exact founding date documented on company site (see Maturity below)

MLSecOps relevance

Security use case: End-to-end risk assessment and runtime protection for multimodal, multilingual GenAI systems, prompt/output inspection, model risk scoring, and shadow AI discovery, across the entire AI lifecycle.

Protected component: Models, LLM/agent applications, and enterprise AI usage (via AI Firewall, Model Scanning, Agent Scanner, AI Red Teaming, and AI Lens/Usage Control products).

Threat coverage: Prompt injection, adversarial inputs, model risk, unauthorized/shadow AI tool usage, and agent-level runtime abuse, across text, image, and video modalities and multiple languages.

Evaluation: https://www.deepkeep.ai — product/platform documentation, with a bookable demo and an agent-scanning trial ("Scan your Agent") linked from the homepage.

Maturity: Holds granted US patent US-12450363-B2 ("Device, system, and method for protecting ML/AI/LLM/DL units"). Covered by Gartner (AI TRiSM, AI Security Platforms, AI Usage Control, AI Security Testing, and Multimodal research), named a Key Innovator in AI Security by Omdia, a Digital Innovator by Intellyx, and a Leader in MarketsandMarkets' AI security market coverage.

Disclosure

• I am affiliated with this resource.
  If affiliated, describe the relationship:

Disclosure: I am affiliated with this project as VP of Marketing.

Technical review

Technical review: Reviewed 2026-08-21. Evidence checked: https://www.deepkeep.ai (platform/product pages), https://www.deepkeep.ai/about (patent US-12450363-B2, analyst recognitions).

Checklist

• I have read and followed the contribution guidelines.
• - [x] The resource addresses a concrete MLSecOps or AI-security problem.
• - [x] The resource is publicly usable or independently evaluable.
• - [x] The description is factual, neutral, and one sentence long.
• - [x] The canonical URL contains no referral or tracking parameters.
• - [x] This PR proposes one resource or one focused change.