Source: www.lastwatchdog.com – Author: bacohido
New York, NY, Apr. 3, 2025 — YRIKKA has released the first publicly available API for agentic red teaming of Visual AI assets. This release comes at the heels of YRIKKA successfully raising its pre-seed funding round of $1.5M led by Focal and Garuda Ventures.
The company was founded by Dr. Kia Khezeli (CEO) and Dr. John Kalantari (CTO), both of whom previously led machine learning projects with the Mayo Clinic, NASA, Intel, and Google. The duo recognized the lack of tools and expertise needed for organizations to build trust in AI systems before and after deployment. YRIKKA’s platform enables decision-makers to assess the reliability and detect vulnerabilities of visual AI assets through automated testing, contextual evaluation, and adversarial red teaming. The platform helps reduce risk and improve AI performance in dynamic environments.
“We saw firsthand how AI models can be powerful yet unpredictable, especially in mission-critical environments,” said Khezeli. “YRIKKA’s goal is to ensure organizations can deploy AI with confidence by providing intuitive tools that help decision-makers understand when and to what extent they can rely on their AI assets.”
Khezeli
YRIKKA’s approach has already attracted high-profile interest. The company has secured multiple contracts with the Department of Defense, including a $1.9M a contract with the U.S. Army to automate the test and evaluation of computer vision models for defense applications. With a growing foothold in various high-stakes sectors, YRIKKA has positioned itself as a key player in AI safety and reliability, particularly for government and enterprise customers concerned with the impactful consequences of deploying AI assets in the real-world.
The company is aiming to address a crucial need in the AI ecosystem: ensuring that AI models are not only functional but also safe, reliable, and resilient to changes in operational context including adversarial threats. The platform serves industries where AI decision-making must be robust, from autonomous defense systems to clinical diagnostics and industrial automation. By integrating synthetic data generation, adversarial red teaming, and fine-tuning capabilities, YRIKKA significantly reduces the time required to validate AI models—from months to minutes.
“AI models – and visual AI models in particular – have the potential to drive massive innovation in large sectors of the economy,” said Arpan Punyani, co-founder and General Partner at Garuda Ventures. “YRIKKA is building a critical suite of tools to help decision makers put these models into production with conviction. We are thrilled to be partnering with such an ambitious team tackling an important problem in AI.”
“As AI moves into the physical world, trust and reliability are critical,” said Daniel Darling, Managing Partner at Focal. “YRIKKA delivers the most rigorous and scalable vision model testing—instant customer adoption proves it.”
About Yrikka: YRIKKA’s mission is to ensure AI deployed in the real world aligns with human values. YRIKKA is the leading startup in visual AI red teaming, building towards a future where humans and AI agents can automatically ensure the reliability of vision models in relevant contexts of use. By enabling rapid, automated assessments tailored to specific operational contexts, YRIKKA helps decision-makers establish trust in AI systems before and after deployment. For more information, please visit https://yrikka.com/.
Media contact: Loopr PR & Marketing, Wahid@looprmarketing.co
The post News alert: YRIKKA’s ‘Red Teaming’ API advances AI safety, reliability in high-stakes applications first appeared on The Last Watchdog.
Original Post URL: https://www.lastwatchdog.com/news-alert-yrikkas-red-teaming-api-advances-ai-safety-reliability-in-high-stakes-applications/
Category & Tags: News Alerts,Top Stories – News Alerts,Top Stories
Views: 1
