Job Title: Product Director, Physical AI Platform Location: Remote (US-based) Reports To: Chief Technology Officer About DeepHow DeepHow is a Physical AI platform for industrial manufacturing, pharmaceuticals, and utilities that helps organizations capture expert know-how, turn it into dynamic work instructions, and drive verified execution on the front line. The platform spans knowledge capture and sharing, AI-powered verification through Smart Compare and photo/video validation, and time and motion intelligence through guided workflows, SOP adherence, and real-time execution visibility. DeepHow supports customers from knowledge capture to verified execution, with strategic account expansion often centered on verification, AI-guided workflows, and time and motion intelligence. Role Description We're looking for a Product Director, Physical AI Platform to lead the build-out of our next-generation Physical AI platform. In this role, you will be the architectural and strategic force behind our platform’s "full layer cake"—from the technical substrate to business outcomes. You will not merely ship capability; you will exercise upstream judgment to determine which problems are actually worth solving, ensuring our platform is understandable, scalable, and delivers durable value to manufacturing enterprises. You’ve worked with AI pipelines and you understand APIs, security, governance, service oriented architecture, ontologies, and scalability at an enterprise level. This is a senior, high-ownership role. You'll lead platform product strategy across DeepHow's core product lines — Knowledge Capture & Transfer, Operational Output Verification, Live SOP Verification, and Time and Motion AI — working closely with engineering, design, partners, and our manufacturing customers to set direction and ship products that matter on the factory floor. What You'll Own Define the Platform Strategy & Roadmap: Act as a product leader who builds and leads platforms that "compress the path from first access to first value". You will own the vision for our Physical AI ecosystem, standardizing where necessary and federating where appropriate to maintain agility. You will define and drive the platform roadmap. Prioritize ruthlessly against customer outcomes, not just feature velocity. Build Technical Foundations: Develop a deep understanding of the platform's technical underpinnings, including AI architecture, data pipelines, and commercially available solutions such as the Nvidia Cosmos stack. You will guide engineering teams on "how things work," ensuring we build robust predictive models (e.g., time series forecasting for manufacturing optimization) that are reliable at enterprise scale. Customer Discovery: In the "End User Era," you will prioritize the needs of internal teams, partners, developer community, and end-users who champion our software. You will know the difference between what customers ask for and what will actually move the needle. You will invert the traditional enterprise sales model by designing a platform where the product itself does the selling. You will identify specific "annoyances"—not just high-level business goals—to solve for builders on our platform, turning their daily frustrations into opportunities for innovation. Cross-functional & Iterative Execution You will partner closely with engineering and design to ship high-quality products. You will write crisp specs and run tight discovery-to-delivery cycles. Balance long-term platform thinking with near-term customer commitments. You will implement a routine of monthly key performance indicators (KPIs) and "in-market" experiments. You will pilot new ideas quickly, leveraging data to decide whether to formalize them or pivot, ensuring we do not invest heavily in unproven hypotheses. Go-to-Market Partnership Support Sales and Customer Success as the domain expert — in enterprise deals, customer QBRs, and product marketing conversations. You can talk to customers’ CIOs, IT, and AI Teams on the same day and land differently with each. AI Product Judgment DeepHow's products are AI-native — from step detection in live video to automated activity classification to computer vision-based output verification. You bring clear thinking on what AI solutions can and can't do reliably in a manufacturing context, and you set appropriate accuracy, latency, and confidence thresholds that customers can actually trust. What You Bring ● An Expert Operator: You bring experience in leading vision, strategy, and execution for enterprise-grade platforms. You have a proven track record of scaling B2B/C platforms across complex industries (e.g., industrial, supply chain, or TMT). You are comfortable operating in environments where AI is infiltrating workflows and changing how organizations function. ● Technically Fluent: Solid understanding of the ML development lifecycle (data prep, training, deployment, monitoring) and MLOps concepts, with prior experience owning a developer-facing or technical product — enough technical depth to earn credibility with ML engineers and data scientists without needing to code. You possess deep knowledge of machine learning paradigms, LLMs and VLMs; and can speak the language of engineers. You understand the trade-offs between technical debt and high-impact feature delivery. ● A Systems Thinker: You understand that the hardest platform problems are often organizational, not just technical. You know how to bring engineering organizations along and how to build platforms that stay "legible at scale". ● Committed to Excellence: You understand that platform product management is not just about building a product; it’s about aligning with customer needs, making solutions “builder-friendly”, and managing developer ecosystems. ● 7+ years of product management experience, at least 3 years with Platform Product Management, and meaningful time spent in industrial software development, data-pipelines, and AI platforms. ● Outcome orientation — you define success in customer terms. ● Strong communication — you write clearly, you can present to a technical team and the C-suite, and you bring alignment rather than confusion ● Bonus: Hands-on experience with computer vision, video AI, or machine learning products in industrial or operational settings The Product Lines You'll Work Across Product What It Does Knowledge Capture & Transfer (KCT) Capture explicit and tacit know-how 90% faster than traditional methods. Reduce onboarding time by up to 77% and create a scalable operation that preserves expertise across sites. Operational Output Verification (OOV) Verify work outputs out of the box with accuracy. Cross-reference photo and video evidence of work outputs against SOPs and documentation to ensure right-first-time execution. Live SOP Verification (LSV) Detect non-conformant steps in real time with up to 96% accuracy. Validate operator steps as they happen and highlight deviations immediately, improving first-pass yield. Time and Motion AI (TMA) Balance lines and complete time & motion analysis faster than current semi-automated approaches. Automatically classify operator activity, eliminate waste, and drive lean manufacturing improvement. Why DeepHow ● The problem is real and urgent — skilled labor shortages and knowledge loss are among the most acute challenges facing manufacturers globally. The work you ship has a direct, visible impact. ● AI at the frontier of physical operations — DeepHow operates at the intersection of computer vision, LLMs, and the shop floor. It's a technically interesting space where accuracy, trust, and usability all matter deeply. ● Customers who care — our users are people who take pride in how things are built. They're not passive software users; they're active partners in making the product better. ● A seat at the table — this is a principal-level role with real influence over product direction, not a ticket-writer role. ● Competitive compensation, equity, and benefits We would love to hear from you.