Senior Technical PM, Applied AI

About the Company: Netomi is the leading agentic AI platform for enterprise customer experience. We work with the largest global brands like Delta Airlines, MetLife, United, and others to enable agentic automation at scale across the entire customer journey. Our no-code platform delivers the fastest time to market, lowest total cost of ownership, and simple, scalable management of AI agents for any CX use case. Backed by WndrCo, Y Combinator, and Index Ventures, we help enterprises drive efficiency, lower costs, and deliver higher quality customer experiences. Want to be part of the AI revolution and transform how the world's largest global brands do business? Join us! Netomi is hiring a Senior Technical PM to own the product strategy and roadmap for our Agents platform. You will decide what we build and why, tie every investment to enterprise customer and business outcomes, and get capabilities shipped and adopted by some of the world's largest brands. This role is also firmly hands-on. You will work directly with AI labs, understand new model capabilities early, build experiments and prototypes yourself, design evaluations and debug failing agents. You use that hands-on work to make better, faster product decisions. Product and business ownership Business focus: tie the roadmap to measurable business outcomes such as autonomous resolution rate, cost per resolution, CSAT, revenue expansion and competitive wins. Know where Netomi wins and loses against competitors, and shape our position. Product strategy and roadmap: own the multi-quarter roadmap for Agents end to end, from problem discovery and prioritization to delivery and outcome tracking. Make and defend hard trade-offs when priorities compete. Customer and partner needs: work directly with enterprise customers, Customer Success, Delivery and Sales to find the highest-value problems. Translate them into clear problem statements, PRDs, user stories, success metrics and dependency maps. Cross-functional leadership: drive alignment across engineering, research, data science, QA and go-to-market on scope, dependencies and launch readiness, without direct authority. Planning and execution: run the cadence of quarterly roadmap inputs, monthly scope lock, weekly execution reviews and launch readiness sign-offs. Launch and adoption: own launch checklists (instrumentation, enablement, success criteria), track adoption per capability and publish a monthly Agent Outcomes readout across deployments. Post-release learning: run monitoring windows and post-incident reviews, and feed what you learn back into the roadmap. Product areas you will own Agent reasoning and orchestration: multi-step task planning, action selection, retry and fallback logic, and supervisor and sub-agent patterns. Tool and API use: reliable, observable, policy-compliant tool calling for CRM, ITSM and custom enterprise systems, including MCP. Memory and context: conversation memory, long-running task state and continuity across channels and sessions. Human-in-the-loop: clarification, graceful escalation to live agents, approval steps for high-risk actions and hybrid autonomy modes. Enterprise governance: guardrails, audit trails, explainable decisions, role-based access, and safe rollout through shadow mode, staged releases and rollback. Hands-on work AI labs and new capabilities: be our product point person with AI labs and model providers. Test new models and features (reasoning, tool use, realtime voice, memory) on our use cases, decide what to build, change or skip, and give labs evidence-based feedback. Experiments and prototypes: vibe code prototypes, build agents, and connect them to real systems through APIs and MCP tools. Run offline tests, shadow mode and canary releases with clear decision criteria, and hand engineering a working reference, not just a document. Evaluation and quality: define the Agent Readiness Scorecard and Go/No-Go criteria. Design evals for task completion, tool success, policy compliance, latency and cost, and keep regression tests running on every model, prompt or tool change. Debugging production failures: read agent traces with engineering, find root causes (wrong action, missed escalation, bad tool call, memory gaps) and make sure each serious failure gets a fix and a test. Model choices: inform model selection and routing decisions by weighing quality, latency and cost with real data. RequirementsProduct skills Roughly 4-8 years of relevant experience, ideally from a deeply technical product role, applied AI engineering or research engineering with clear product ownership. Proven ownership of a complex technical product area, with features delivered to enterprise customers and measurable business results. Strong product judgment: you can prioritize, work with customers and get capabilities shipped. Customer-first: you turn enterprise needs into clear, actionable requirements. Excellent communicator who can align engineering, research and go-to-market teams and present clearly to executives. Hands-on skills Comfortable vibe coding, building agents, connecting APIs and MCP tools, designing evaluations and debugging production failures. Deep understanding of LLM reasoning, tool calling, multi-step task execution, orchestration patterns, evaluation methods and production observability for agents. Technical fluency with APIs, cloud environments and AI deployment workflows. Nice to have Experience shipping AI agent products at a conversational AI or enterprise AI platform company. Direct experience working with AI labs or model providers (early-access programs, model evaluations). Familiarity with agent frameworks and protocols (e.g., LangGraph, MCP) and with voice AI. Experience in regulated or mission-critical industries (airlines, banking, healthcare, payments) and with large SI partners. Public work you can show: a repo, demo, blog post or talk. Netomi is an equal opportunity employer committed to diversity in the workplace. We evaluate qualified applicants without regard to race, color, religion, sex, sexual orientation, disability, veteran status, and other protected characteristics. #J-18808-Ljbffr