Senior Applied AI Product Manager

Job Description QAD AI is building the Persona Agent platform the product that lets manufacturers deploy AI agents across procurement, sourcing, order management and planning. As we scale deployments through our Forward Deployed Engineering motion, the platform has to mature just as fast as the field learns. That is this role. The Senior Applied AI Product Manager owns the Persona Agent platform roadmap end-to-end. You decide what the platform does out-of-the-box, what becomes reusable, and what stays a one-off turning the signal coming back from real customer deployments into a coherent, defensible product. You report directly to the Head of AI and are the product counterpart to the field. What Youll Own 1. The Persona Agent platform roadmap You own the roadmap for the platform and its agent capabilities the vision, the priorities, and the trade-offs. You decide what gets built, in what order, and why, and you hold the line on it. Vision strategy: Set and maintain the platform roadmap against a clear thesis of where agentic manufacturing software is going. Prioritisation: Own the backlog, prioritisation and sequencing across competing customer and internal demands. Platform vs. custom: Decide what belongs in the platform vs. what stays a customer-specific build the single most important call you make repeatedly. Agent scope: Define what a Persona Agent should do out-of-the-box for each process (P2P, O2C, S2C, P2M) and where the boundaries sit. 2. The field-to-product loop The FDE team and AI Solutions Architects are deploying agents inside real customers every week, and each engagement surfaces patterns, gaps and hard-won solutions. You are the person who catches that signal and decides what becomes product. This loop is the engine of the platform without it, the roadmap is guesswork; with it, every deployment makes the product better. Capture: Run a structured intake of field signal product gaps, recurring exception logic, integration patterns, autonomy models from the FDE team and Architects. Triage: Decide which field solutions are one-offs and which are patterns worth productising into reusable platform capabilities. Productise: Turn recurring field builds (e.g. EDI error handling, chat-driven PO creation, per-category autonomy dials) into first-class product features. Communicate: Close the loop back to the field so Architects and FDEs know whats coming and can design against it. 3. Delivery with engineering You work hand-in-hand with the AI engineering and Applied AI teams to ship the roadmap writing crisp specs, making scope calls, and keeping delivery honest against outcomes rather than output. Specs: Write clear product specs and acceptance criteria that engineers can build against without ambiguity. Metrics: Define how each capabilitys success is measured adoption, agent performance, deployment-time saved and track it. Trade-offs: Make the hard scope trade-offs in the moment, and own the consequences. 4. Evaluation quality of agent behaviour Agent products live or die on whether the agent behaves. You own the product view of agent quality defining what good looks like, and making sure evaluation is built into the platform, not bolted on. Evals: Define the evaluation criteria for agent behaviour per process and per capability. Instrumentation: Ensure attribution and telemetry are product requirements, so every deployment can prove impact. Trust: Set the guardrails and autonomy defaults that keep agents safe and trustworthy in production. How You Work With the Rest of the Org You sit in the AI organisation reporting to the Head of AI, at the centre of the product, engineering and field triangle. AI Solutions Architects - Primary source of field requirements and validation; you turn their designs into repeatable product. FDE team (Services) - Your richest signal on what recurs across builds; you productise their accelerators. Applied AI / Engineering - Your build partners; you spec, prioritise and ship the roadmap with them. Head of AI - Your manager and the roadmaps executive owner; you bring the plan, the trade-offs and the field evidence. Qualifications What Youll Bring Must-have 5 10 years in product management, with a track record owning a technical platform or developer-facing / enterprise product roadmap end-to-end. Technical fluency with modern AI you understand agents, LLMs, tool use, RAG, evals and autonomy well enough to make sound product calls and earn engineers trust, without needing to build it yourself. Demonstrated ability to turn messy, real-world signal into a coherent roadmap and to say no to most of it. Strong sense for the platform-vs-custom boundary: what to generalise, what to leave bespoke, and how to avoid a roadmap of one-offs. Excellent written communication crisp specs, clear prioritisation rationale, and roadmaps stakeholders can act on. Comfort operating close to customers and field teams; you make decisions on evidence, not opinion. Strongly preferred Experience with enterprise software ERP, supply chain, or manufacturing or the ability to get fluent in P2P, O2C, S2C and P2M quickly. Experience shipping AI/ML or agentic products, and familiarity with evaluation frameworks for model or agent quality. Prior work bridging a field/forward-deployed team and a core product org. Nice-to-have A technical background (engineering, data, or applied science) earlier in your career. Familiarity with cloud platforms (AWS), APIs and integration patterns common to enterprise deployments. What Success Looks Like - First 90 days: A clear, evidence-based platform roadmap in place, built from current field signal, with buy-in from the Head of AI, Architects and engineering. - 6 months: Recurring field builds productised into platform capabilities; a working field-to-product intake running on a regular cadence. - 12 months: Measurable reduction in per-deployment build effort as the platform absorbs what was once custom; agent quality and adoption trending up on defined metrics. The bar You are the person who makes sure the platform gets smarter every time the field ships something so QAD is building a product, not a backlog of bespoke deployments. Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.