Problem

P-1 AI is tackling the shortage of engineering bandwidth in industries that design complex physical systems. Physical-product AI is difficult to build because the relevant training data is scarce: there are relatively few unique designs, and the available designs are often proprietary and lack cohesive multi-physics models. At the same time, human engineers spend substantial time on repetitive cognitive work, creating a capacity, cycle-time, and talent-shortage problem rather than a lack of existing engineering software.

The initial killer use case is data-center engineering, especially cooling and critical-power systems. In this workflow, engineers must synthesize requirements, select components, perform design trade studies, create design artifacts, and run power or thermal analyses. P-1 AI’s demonstration completed this end-to-end work in 23 minutes versus roughly seven to ten working days for a human engineer, illustrating the potential economic value of compressing engineering cycles and expanding team capacity.

Product / Service

P-1 AI is building Archie, an agentic AI engineer for the physical world. Archie initially operates at the level of a junior mechanical and electrical engineer, with quantitative and spatial reasoning over product-design domains. Its technology combines a custom agentic harness, structured design representations, continual skills learning, custom post-trained models, and physics-based, supply-chain-informed synthetic data. Rather than replacing CAD, PLM, analysis, or productivity software, Archie uses the same tools as human engineers, including SolidWorks, AutoCAD Electrical, Excel, SharePoint, PLM systems, and specialized thermal and power-analysis software.

The delivery model is an AI teammate embedded in an existing engineering organization, designed to feel like an outsourced junior engineer joining the team. Archie follows the customer’s existing processes and produces work products ready for human review, while taking on repetitive, time-consuming tasks and working continuously across shifts. The intended benefit is to free senior engineers for systems thinking, architecture, innovation, and customer engagement while giving industrial OEMs more flexible engineering capacity.

Market

P-1 AI sits at the intersection of industrial AI, engineering-automation software, and engineering AGI for physical systems. It is initially focused on data-center cooling and critical-power applications, with expansion planned across automotive, aerospace and defense, industrial systems, and other hardware domains. The company’s materials do not identify a named direct competitor; instead, they explicitly position Archie as complementary to existing engineering tools, with differentiation centered on cognitive automation and domain-specific physical reasoning.

The company is not presented as pre-revenue, but its public materials do not disclose revenue or pricing. It has reported several design partnerships with leading industrial OEMs, demonstrated end-to-end workflows, and showcased data-center design capabilities with Daikin Applied Americas and in support of the NVIDIA DSX reference design. P-1 AI announced an initial $50 million Series A financing led by NEA in July 2026, following a $23 million seed round led by Radical Ventures in 2025, and described the Series A as reflecting strong product execution and commercial traction.