应用与代理AI工程师
Applied & Agentic AI Engineer
加入Sedgwick，你将参与真正有意义的事业。我们的33,000名同事每天为全球面临意外的人们提供帮助。我们邀请你与我们共同成长，体验我们的关怀文化，享受工作与生活的平衡。在这里，你所能实现的没有界限。
Newsweek 将 Sedgwick 评为美国最佳工作场所全国顶级公司
被认证为卓越工作场所®
财富金融与保险行业最佳工作场所
应用与代理 AI 工程师 职位职责
- 设计并部署基于大语言模型（LLM）和代理的AI解决方案，以转变理赔受理、保单解读、欺诈检测和解决流程。
- 设计端到端检索增强生成（RAG）系统，利用企业知识库、保单文件、标准操作程序（SOP）和历史理赔数据。
- 构建能够推理、规划和执行多步骤理赔流程的自主和半自主代理。
- 开发状态感知的工作流编排层，管理交互中的上下文、记忆和任务顺序。
- 实现计划和反思循环，将复杂的理赔场景分解为结构化子任务。
- 通过函数调用和与理赔系统、客户关系管理（CRM）平台、文档存储库和分析工具的安全API集成，实现动态工具使用。
- 使用大语言模型（LLM）开发文档智能处理流程，包括摘要、实体提取、分类、验证和时间线重建。
- 设计结构化提示框架，确保确定性输出和领域感知推理。
- 构建多代理系统，协调文档审查、保障分析、合规检查和决策支持。
- 实现人机协作检查点，用于AI驱动决策的升级、审核和覆盖。
- 开发护栏、输出验证层和幻觉缓解策略。
- 使用模式、类型验证和确定性后处理逻辑来强制结构化输出。
- 优化令牌消耗、推理延迟和云基础设施成本。
- 使用容器化和云原生架构部署可扩展的AI微服务。
- 实现对模型漂移、检索质量下降、推理失败和工作流中断的监控。
- 维护详细的模型决策、代理推理步骤和工具执行的审计日志。
- 开发评估框架，测试推理准确性、工作流效率和系统稳定性。
查看英文原文
By joining Sedgwick, you'll be part of something truly meaningful. It’s what our 33,000 colleagues do every day for people around the world who are facing the unexpected. We invite you to grow your career with us, experience our caring culture, and enjoy work-life balance. Here, there’s no limit to what you can achieve.
Newsweek Recognizes Sedgwick as America’s Greatest Workplaces National Top Companies
Certified as a Great Place to Work®
Fortune Best Workplaces in Financial Services & Insurance
Applied & Agentic AI EngineerJob Responsibilities
- Architect and deploy LLM-powered and agentic AI solutions that transform claims intake, policy interpretation, fraud detection, and resolution workflows.
- Design end-to-end retrieval-augmented generation (RAG) systems leveraging enterprise knowledge bases, policy documents, SOPs, and historical claims data.
- Build autonomous and semi-autonomous agents capable of reasoning, planning, and executing multi-step claims processes.
- Develop stateful workflow orchestration layers that manage context, memory, and task sequencing across interactions.
- Implement planning and reflection loops that decompose complex claims scenarios into structured subtasks.
- Enable dynamic tool use through function calling and secure API integrations with claims systems, CRM platforms, document repositories, and analytics tools.
- Develop document intelligence pipelines using LLMs for summarization, entity extraction, classification, validation, and timeline reconstruction.
- Design structured prompt frameworks that enforce deterministic outputs and domain-aware reasoning.
- Build multi-agent systems that coordinate document review, coverage analysis, compliance checks, and decision support.
- Implement human-in-the-loop checkpoints for escalation, review, and override of AI-driven decisions.
- Develop guardrails, output validation layers, and hallucination mitigation strategies.
- Enforce structured outputs using schemas, type validation, and deterministic post-processing logic.
- Optimize token consumption, inference latency, and cloud infrastructure costs.
- Deploy scalable AI microservices using containerization and cloud-native architectures.
- Implement monitoring for model drift, retrieval quality degradation, reasoning failures, and workflow breakdowns.
- Maintain detailed audit logs of model decisions, agent reasoning steps, and tool executions.
- Develop evaluation frameworks to test reasoning accuracy, workflow completion rates, and system reliability.
- Collaborate with data engineering to build embedding pipelines, feature stores, and vector indexing strategies.
- Ensure compliance with Responsible AI standards, data privacy regulations, and enterprise governance policies.
- Partner with claims operations leadership to embed AI capabilities directly into adjuster and supervisor workflows.
- Measure business impact through cycle-time reduction, automation coverage, fraud detection lift, and operational efficiency gains.
Qualifications
- Bachelor’s or Master’s degree in Computer Science, Artificial Intelligence, Engineering, or related field.
- 5+ years of experience building production-grade AI or advanced software systems.
- 2–4+ years of hands-on experience with LLM-powered applications and orchestration layers.
- Strong expertise in retrieval-augmented generation architectures and vector search systems.
- Experience designing and implementing multi-agent systems and workflow orchestration engines.
- Deep understanding of planning loops, contextual memory, and tool-augmented LLM reasoning.
- Strong proficiency in Python and API-driven system design.
- Experience integrating enterprise platforms and building secure connectors.
- Familiarity with Azure OpenAI or similar enterprise LLM environments.
- Experience deploying containerized services and managing CI/CD pipelines.
- Understanding of distributed systems, microservices, and event-driven architectures.
- Experience implementing guardrails, access controls, and auditability mechanisms.
- Strong knowledge of evaluation methodologies for LLM reliability and agent performance.
- Experience in insurance, claims, healthcare, or other regulated industries preferred.
- Ability to translate complex operational workflows into scalable, AI-driven autonomous systems.
Sedgwick is an Equal Opportunity Employer and a Drug-Free Workplace.
If you're excited about this role but your experience doesn't align perfectly with every qualification in the job description, consider applying for it anyway! Sedgwick is building a diverse, equitable, and inclusive workplace and recognizes that each person possesses a unique combination of skills, knowledge, and experience. You may be just the right candidate for this or other roles.
Originally posted on Himalayas