Senior Product Manager, Agentic Application

Duck Creek Technologies•Bay Area, CA

9d•$132,000 - $235,200•Hybrid

About The Position

Duck Creek Technologies is seeking a Senior Product Manager, Agentic Application to own the product strategy, roadmap, and execution for an agentic AI application. This application automates insurance product configuration for Duck Creek carriers by enabling AI agents to understand complex carrier product documents and translate them into structured, validated configurations. This role bridges the gap between agentic AI, insurance product knowledge, and enterprise software delivery, requiring independent judgment and collaboration across various teams including Product, Engineering, Architecture, Data Science, AI Platform, Customer Success, Professional Services, and carrier stakeholders. The goal is to deliver governed, accurate, and scalable agentic configuration capabilities.

Requirements

Bachelor's degree or equivalent experience in Computer Science, Information Systems, Engineering, Data Science, Business, Insurance, Product Management, or a related field.
6+ years of experience in technical product management, AI product management, platform services, software engineering, or related roles.
Experience owning complex AI, data, platform, or document-processing product capabilities.
Experience working with cross-functional stakeholders, including engineering, architecture, data science, customer-facing teams, and business stakeholders.
Experience translating ambiguous business and technical problems into product strategy, roadmap priorities, requirements, release plans, and measurable outcomes.
Technical product management practices for complex AI, data pipeline, or document processing capabilities.
AI agent design, including LLM-based extraction, deterministic agent patterns, orchestration frameworks, and multi-stage pipeline architecture.
Extraction accuracy concepts, including eval harness design, scoring methodology, baseline comparison, and prompt iteration.
Knowledge graph concepts, including entity relationship modeling, semantic abstraction, dependency graph traversal, and change impact analysis.
Human-in-the-loop design patterns, including review gate placement, correction capture, audit logging, and reviewer experience.
Document understanding across complex insurance documents, including rate manuals, carrier filing documents, underwriting guidelines, and product manuscripts.
Event-driven system design, including versioned event contracts, Pub/Sub patterns, durable state management, and agent checkpointing.
Responsible AI practices, including hallucination detection, deterministic output validation, schema-enforced output gates, and accuracy-gated deployment.
Cloud-native platforms, multi-tenant SaaS architecture, and enterprise security patterns relevant to carrier data handling.
Translating ambiguous carrier configuration needs and AI accuracy requirements into structured product requirements, clear priorities, and measurable outcomes.
Evaluating trade-offs across extraction accuracy, reviewer burden, LOB scalability, infrastructure cost, carrier data privacy, and time-to-market.
Analyzing accuracy, reliability, pipeline performance, and business impact metrics to inform product decisions and prioritization.
Ability to influence stakeholders without direct authority across a broad cross-functional organization, including Product, Engineering, Architecture, Data Science, Customer Success, Professional Services, and carrier customers.
Ability to explain complex AI extraction, knowledge graph, and validation decisions in clear business and carrier terms.
Ability to use AI-enabled tools to support product discovery, analysis, prototyping, documentation, and iterative product development.
Legally authorized to work in the country of the job location.

Nice To Haves

Master's degree or equivalent advanced experience in a relevant field.
Experience with AI agent frameworks, LLM evaluation, knowledge graphs, document understanding, structured data extraction, or agentic AI pipeline design.
Experience in property and casualty insurance, InsurTech, financial services, or another complex regulated industry, particularly in product configuration, underwriting, policy administration, or manuscript management.
Familiarity with event-driven architectures, cloud-native platforms, LangGraph or similar agent orchestration frameworks, and human-in-the-loop system design.

Responsibilities

Define and maintain the product strategy and roadmap for an agentic AI application that automates insurance product configuration, including document ingestion, structured data extraction, knowledge graph development, configuration generation, output validation, and human review workflow design.
Translate carrier product configuration needs, document change workflows, accuracy requirements, and line-of-business expansion goals into clearly defined epics, features, use cases, requirements, user stories, and measurable product outcomes.
Partner with Engineering and Architecture to define scalable, reliable, and maintainable agent pipeline patterns, including agent orchestration design, event-driven communication, durable state management, and cloud infrastructure requirements.
Define product capabilities for a knowledge graph that represents relationships between insurance product elements, enabling intelligent change impact detection, cross-element dependency tracking, and semantic abstraction over raw product document structure.
Drive an eval-first product culture: define accuracy targets by capability, establish a continuous evaluation methodology using AI observability tooling, and ensure no feature ships without a measurable accuracy signal.
Partner with AI Platform, Semantic Insurance Data, Data Orchestration, Security, and Governance teams to support trusted context, responsible AI controls, and reliable execution across carrier document types and product lines.
Define carrier-facing product capabilities, including configuration request intake, pipeline status visibility, reviewer experience, extraction result review, and output approval workflows.
Evaluate trade-offs across extraction accuracy, pipeline reliability, reviewer burden, scalability, carrier data privacy, infrastructure cost, and time-to-market when shaping product priorities and requirements.
Define adoption, accuracy, reliability, and business impact measures — including extraction accuracy, configuration validity rates, human review approval rates, pipeline throughput, LOB onboarding velocity, and carrier time-to-configuration.
Support release planning, readiness, and communication by clarifying scope, dependencies, risks, acceptance criteria, and expected accuracy outcomes.
Maintain awareness of emerging agentic AI frameworks, LLM evaluation methodologies, knowledge graph technologies, document understanding models, and insurance product configuration patterns to inform product direction.