Senior Product Manager

Job Description

This role owns the full product lifecycle for QAD's Copilot, Search, and Conversational Analytics capabilities — from architecture through to market launch. You will define what gets built, why it matters to manufacturing users, and how it reaches them: shaping the product vision, driving engineering delivery, and partnering with GTM to ensure adoption.

The intelligence that powers these surfaces — how queries are understood, how manufacturing context is assembled, how data is retrieved, and how responses are generated — is where QAD's AI moat is built. You will need to engage deeply with these layers, not as an engineer but as the product owner who defines the contracts, quality standards, and sequencing decisions that determine whether they work in production at enterprise scale.

You will report to the Head of Platform Product, with day-to-day direction from the Director of the AI Platform org. This role is evaluated through the quality of your product thinking, your influence on engineering direction, and the outcomes you drive in the market.

The Opportunity

Manufacturing users today cannot query their own operations without analysts, BI tools, or pre-built reports. QAD's AI platform changes this — but only if the Copilot and Search layer is built correctly and lands with users. You will own both sides of that equation.

• Define the semantic search and conversational analytics product that allows manufacturing users to query operational data — orders, inventory, suppliers, quality records — in natural language, without SQL or BI expertise

• Own the intelligence contract: how queries are understood, how manufacturing context is assembled, how retrieval is orchestrated, and how responses are grounded in governed data — the foundational decisions that determine product quality at scale

• Drive the Copilot from concept to customer — including the contextual layer architecture, the grounding contract against the platform's Semantic Layer, and the GTM motion that gets it adopted in manufacturing workflows

• Build the feedback loops that make the product smarter over time: how recurring query patterns surface ontology gaps, how session analytics drive prioritisation, and how discovery findings translate into product improvements

• Establish QAD's conversational analytics presence in market — working with GTM to define positioning, enablement, and the narrative that differentiates QAD's intelligence layer from horizontal AI tools

Key Responsibilities

Copilot & Conversational Analytics

• Own the product definition for QAD's Copilot: how queries are understood, context assembled, data retrieved, and responses generated — specifying the contracts that engineering builds against

• Define the grounding contract: the rules that ensure every Copilot response cites governed manufacturing data, not hallucinated inference — including confidence signalling and graceful fallback behaviour

• Drive the conversational analytics strategy: how natural language queries translate into analytical results across manufacturing data domains without exposing SQL or BI complexity to the user

• Specify the disambiguation model: how the Copilot handles ambiguous queries, missing context, and conflicting data signals in real manufacturing workflows

Search

• Own the product definition for QAD's semantic search layer: indexing strategy, query understanding, entity recognition, ranking, and result structure across manufacturing data domains

• Define the search intent taxonomy — operational (find order, trace shipment), analytical (show trends, compare periods), and diagnostic (why is this delayed, what caused this variance) — and specify result formats by intent type

• Drive the federated search architecture: how results from ERP operational data, analytical stores, and the knowledge graph are ranked and merged into a coherent, useful response

Discovery, Delivery & GTM

• Maintain a structured customer discovery programme — extracting real manufacturing query patterns and translating findings into specific product and architecture implications

• Own the product backlog: writing stories to Staff Engineer level of specificity — API contracts, state machines, data flows, edge cases — that engineering can build without verbal clarification

• Partner with GTM to define the launch and adoption strategy for Copilot and Search: positioning, sales enablement, onboarding playbooks, and the narrative that lands with manufacturing buyers

• Define the quality framework for these surfaces: what signals matter for product performance — retrieval relevance, grounding rate, latency distribution — and how they inform roadmap decisions

Stakeholder Management & Managing Up

• Proactively surface dependencies, risks, and scope changes to the Director and Head of Platform Product — with a resolution proposal, not just an escalation

• Communicate product and architecture decisions clearly to non-technical stakeholders — translating trade-offs into business implications without jargon

• Build credibility with Engineering, Architecture, and GTM through the quality and precision of written work

Qualifications

Product Experience

• 7–12 years in product management, with significant time owning AI-powered, search, or analytics products in enterprise B2B environments — from definition through production launch

• Proven track record owning the full product lifecycle: customer discovery, specification, engineering delivery, and GTM — not just one slice

• Experience working in matrixed organisations — driving outcomes across Platform, Engineering, Architecture, and GTM without direct authority

• Evidence of structured discovery practice applied to a technical product surface — ability to extract product implications from user research, not just UX insights

Core Technical Depth — Required

These are the primary technical domains this role owns. Strong command is expected from day one.

• Semantic search architecture: indexing strategy, query understanding, entity recognition, ranking models, and relevance evaluation at enterprise scale

• Conversational AI product design: how natural language queries are translated into structured analytical results across complex, domain-specific data without exposing backend complexity to users

• LLM-integrated product specification: grounding strategies, hallucination mitigation, response quality gates, and context window management — sufficient to define product contracts engineering builds against

• Data platform fluency: OLAP vs operational DB reads, federated search across heterogeneous stores, and how data freshness constraints affect product quality — sufficient to hold authoritative conversations with data engineers

• SQL and data modelling: sufficient to validate query plans, understand schema implications, and reason through retrieval architecture trade-offs

Working Familiarity — Expected to Grow Into

These domains sit in adjacent platform layers that Copilot and Search consume. Deep expertise is not required on day one, but working fluency is expected — and will expand as the platform scales toward agentic capabilities.

• Intent detection and query understanding pipelines: how user inputs are parsed, classified, and routed across platform layers

• Context assembly and packaging: how metric definitions, entity relationships, and business rules are structured and passed downstream at inference time

• Session and state management: conversation continuity, context windowing across multi-turn interactions, and memory summarisation patterns

• Human-in-the-loop design: approval workflows, confidence thresholds, and escalation paths where AI-generated outputs require human validation before action

Competency Expectations

This role is calibrated to a Staff PM equivalent. Across QAD's four PM competency pillars:

…