卖什么
AI agent that optimizes inference for embedded compute like Jetson
商业模式拆解
买家：谁掌握预算？
未核实ML/perception engineering leads at autonomy and robotics companies running on NVIDIA Orin and Thor.
触发点：什么让他们现在购买？
Standard compilers and libraries do not hit latency targets for autonomy/perception stacks (dynamic workloads, multi-model pipelines, unusual ops).
工作流：什么输入变成什么输出？
Customer coding agent prepares a Request Package from internal model stack (no weights or training data), uploads to RunLocal Platform; its agentic ML compiler optimizes kernels, quantization, fusion, scheduling for Jetson Orin/Thor; customer downloads Deliverable Package and validates locally.
经济账：省了什么，怎么收费？
Value claimed: lower latency, larger models on same compute, cheaper onboard compute, fewer scarce ML compiler hires; pilot-based engagement (1-week pilot offered).
渠道：怎么接触到客户？
Sales-led: Book A Call, 1-week pilot, docs and sign-in platform.
融资历史与报道
| 融资日期 | 轮次 | 金额 | 投资方 | 来源 |
|---|---|---|---|---|
| — | YC 2024 夏季 | — | Y Combinator | ycombinator.com |
档案汇总自公开报道和公司官网。商业模式部分是 MaxBC 对这些材料的解读，不是已确认的事实；未知之处留空，不作猜测。 更新于 2026-10-12.