Text-to-CAD: AI Revolutionizing Hardware Design with Jordan Noone of Zoo

Company Overview & Market Opportunity

• Mission: Zoo is building the world's first cloud-based, GPU-optimized computational geometry engine to bring AI automation to the hardware development lifecycle.
• Market Size: Targets a ~$1 trillion market comprising labor and software license costs associated with global hardware development (ranging from furniture to aerospace).
• Problem Statement: The industry has remained largely unchanged since the 1980s, relying on labor-intensive workflows where engineers manually click mice to define every geometric line, circle, and surface.
• Target Segment: Focuses on companies at the intersection of extreme software and extreme hardware development (e.g., robotics, autonomous vehicles, aerospace) where software automation exists but hardware design remains a bottleneck.

Product & Technology Stack

• Core Engine: Developed a proprietary computational geometry engine from scratch using CUDA (GPU-level machine language), a skill set largely lost since the 1980s, enabling code-based access to geometric data.
• Modeling App: A CAD interface allowing engineers to edit parts via mouse, code (scripting, equations, logic), or AI prompts, replacing the traditional "mouse-only" workflow.
• Text-to-CAD: A generative AI product launched in December (previous year) that functions as a ChatGPT interface for hardware, generating CAD files from text prompts.
  • Launch Metrics: Generated 13,000 CAD files on the first day of launch, significantly exceeding inbound infrastructure expectations.
• Workflow Integration: Upcoming V1.0 release will merge the Modeling App and Text-to-CAD, allowing users to generate designs via AI, then refine them manually or via code within a single interface.
• API & Open Source: Offers open-source tools to drive API traffic and ecosystem development while maintaining proprietary defenses in the core engine; APIs allow third parties to build bespoke workflows (e.g., manufacturing-aware tools).

Strategic Roadmap & Future Developments

• Manufacturing Integration: Expanding beyond design into downstream manufacturing by adding CNC tool path generation and Design for Manufacturability (DFM) checks directly into the ML models.
• End-to-End Testing: Building an in-house factory in Inglewood, CA (previously Relativity Space HQ) to perform end-to-end verification of toolpaths by running physical CNC tests on generated code.
• Infrastructure Growth: Doubling down on Machine Learning (ML) and computational geometry teams; constructing a first local data center to house GPUs for both geometric rendering and ML training to reduce costs and ensure scalability.
• Team Expansion: Currently ~20 employees; plans to double to ~50 staff by the end of the next year, focusing on hiring for ML, infrastructure, and mechanical engineering roles.

Leadership & Team Background

• Jordan Noon (Executive Chairman): Aerospace engineer who was the youngest person to get FAA clearance to fly a rocket (age 19); co-founded Relativity Space (founded at age 22, valued at $4.2B); previously worked at SpaceX, HRL, and Blue Origin.
• Jess (CEO): Former early program manager at Docker; served as Chief Product Officer at Oxide Computer Company; brings deep software infrastructure and hardware automation experience.
• Team Composition: Unique blend of computational geometry engineers (a rare skill set), ML engineers, infrastructure engineers, and mechanical engineers for usability testing.

Funding & Financials

• Total Raised: Approximately $6 million in private venture capital to date.
• Investor Composition:
  • Pre-seed: Led by Embedded Ventures (a fund co-founded by Jordan Noon and partner Jenna).
  • Seed (End of 2022): Led by UK-based Venrex (strategic portfolio partners); included angel investments from GitHub co-founder Tom Preston-Werner and former CEO Nat Friedman.
• Operational Model: Described as a "very lean, very efficient" team that bootstrapped product development before significant external capital injection.

Key Disagreements & Industry Shifts

• Industry Stagnation: No engineering software IPOs have occurred since 1986; current solutions are merely "porting" legacy software to the browser without changing the core workflow or enabling automation.
• Data Accessibility: The fundamental bottleneck in hardware AI is the inability to "talk" to geometry data with code, requiring interfaces that rely on mouse clicks rather than APIs.
• Workflow Asymmetry: Modern software development utilizes continuous integration and automated testing, whereas hardware development remains stuck in manual, siloed processes that cause team bloat and interface errors.