
The data, models, and platform to fine-tune edge-optimized SLMs.
Product: Task-specific language models (TLMs/GLiNER) and a Personalization API
Performance focus: CPU/edge-first, low-latency inference (claims of <50ms for some models)
Headquarters: Palo Alto, California
Team size: ~21 employees
Funding: ~$24–25M across Nov 2024 pre-seed and May 2025 seed
Deploying efficient, task-optimized language models for low-latency inference on CPUs and edge devices for enterprise extraction, classification, and personalization tasks.
Artificial intelligence / Machine learning
$7.0M
$17.5M
“Backed by institutional investors including Khosla Ventures, Insight Partners, and M12; participation from additional angels and funds”
| Company |
|---|