Portfolio Spotlight: How NeuralSpace is Redefining Edge AI
By Olivent Team
When we first met the NeuralSpace founding team in late 2024, they were three researchers from the Technion with a bold thesis: the future of AI isn't in the cloud, it's at the edge.
The Problem
As AI models grow more powerful, the latency and bandwidth costs of cloud inference become prohibitive for many applications. Autonomous vehicles, industrial robotics, and real-time medical devices all need AI that runs locally, not in a data center thousands of kilometers away.
The NeuralSpace Solution
NeuralSpace has built a hardware-software co-design platform that optimizes neural network architectures for specific edge deployment targets. Their compiler takes standard models and produces highly optimized versions that run 10-50x faster on commodity hardware.
Key Metrics
| Metric | Value |
|---|---|
| Model compression | Up to 50x |
| Latency reduction | 10-50x vs cloud |
| Power efficiency | 5x improvement |
| Supported hardware | 12+ chip families |
Why We Invested
Three things stood out:
- World-class team - Deep expertise in compiler design and neural architecture search
- Massive market - Edge AI is projected to be a $50B+ market by 2030
- Strong early traction - Three Fortune 500 pilot programs before raising their seed round
What's Next
NeuralSpace is currently scaling their engineering team and preparing for their Series A. We're excited to continue supporting them as they bring edge AI to industries that need it most.
Olivent is an investor in NeuralSpace. This post is for informational purposes only.