Given available information, the best-fit ICP looks like:
- Company size: Mid-market to enterprise hardware makers and product teams (mid-sized to large organizations) that ship devices at volume.
- Industries: Consumer electronics, industrial IoT, security/monitoring, robotics, defense-adjacent dual-use.
- Roles: Embedded ML engineers, CV/ML leads, product engineering, hardware platform teams.
- Primary pain points:
- Need low-latency inference without cloud dependence
- Power/compute constraints
- Need reliable detection/tracking on-device