EdgeRunner AI, a new company building Generative AI for the edge, recently announced its official launch and the closing of a $5.5 million seed funding round led by Four Rivers Group with participation from Madrona Ventures and strategic angels.
This funding round will go towards building EdgeRunner AI’s innovative platform, which is powered by state-of-the-art transparent models designed to operate seamlessly on any device or hardware without requiring internet access. And this technology will transform how enterprises and organizations responsibly deploy Generative AI, providing them with hyper-personalization and domain specific use cases.
By creating small task-specific open models optimized for any chip, EdgeRunner AI elevates SLMs (Small Language Models) into Ultra-Efficient Language Models (UELMs). This enables organizations to run AI models locally on any device or hardware, resulting in improved performance, increased data privacy, near zero latency, and reduced power consumption.
KEY QUOTES:
“The challenge with Generative AI today lies in its broad applicability without addressing specific business needs. At EdgeRunner, we empower organizations to own their AI. From my military experience in Eastern Europe, I saw the need for personalized, domain specific, and secure AI in isolated environments. Black box proprietary models are aligned to the cloud rather than the user. Understanding the biases within these opaque models is complex and critical for both usability and safety. Our approach ensures full transparency of models from weights and code to architecture and datasets, that doesn’t require you to move your data and can run on any device and chip.”
- Tyler Xuan Saltsman, Co-Founder and CEO of EdgeRunner AI
“Enterprises and governments don’t need GPT-5 or AGI. What they need is to responsibly deploy Generative AI to solve their real-world business challenges using multiple small, task-specific open models working together to create swarm intelligence.”
- Colton Malkerson, Co-Founder & COO of EdgeRunner AI
