Agentic AI Forward Deployed Engineer

Experience: 8-12 Years Team: Lead a pod of 3 Agent Engineers Reporting To: Delivery Manager / Engineering Manager We are looking for an Agentic Forward Deployed Engineer to work directly with business stakeholders, rapidly translate complex challenges into production-ready agentic solutions, and lead the delivery of autonomous and multi-agent systems at enterprise scale. In this role, you will operate at the front line of delivery, embedding closely with clients to design, build, deploy, evaluate, and continuously improve Business Transformation Agents that automate and reimagine processes such as procurement approvals, invoice dispute resolution, onboarding, claims processing, compliance workflows, and other critical business operations. What You'll Do Partner with stakeholders to quickly conceptualize business processes as agentic solutions and deliver working prototypes in days, not weeks.Design and build single-agent and multi-agent systems in Python using leading Agent Development Kits (ADKs).Drive operational efficiency through reduced cycle times, lower manual effort, improved accuracy, and lower cost-to-serve.Engineer production-grade agents using:Prompt EngineeringContext EngineeringPrompt CachingRAG / Context Graph RetrievalMemory ManagementTool & Function CallingModel Context Protocol (MCP)Multi-Agent OrchestrationIntegrate agents with enterprise systems using standards-based APIs, secure authentication, and scalable integration patterns.Create reusable agent components, skills, templates, and tool libraries that improve delivery speed and consistency.Implement eval-driven development through testing frameworks and evaluation harnesses that measure correctness, safety, quality, and reliability.Own AgentOps and DevSecOps practices including CI/CD, observability, telemetry, security, governance, and Responsible AI controls.Continuously improve agents using production telemetry, evaluation results, and stakeholder feedback.Stay current with emerging agent frameworks and engineering patterns, bringing new ideas and best practices into delivery.Lead and mentor a pod of 3 Agent Engineers while helping shape the future of Agentic AI delivery. Technology Environment Programming Language Python (preferred)Agent Development Kits Google ADKLangGraphCrewAIOpenAI Agents SDKAWS Bedrock AgentCoreMicrosoft Agent Framework / Semantic KernelLLMs Azure OpenAIClaude (via Bedrock)GeminiOther enterprise LLMs selected based on quality, latency, and costIntegration MCP (Model Context Protocol)APIs & MicroservicesSecure Authentication & AuthorizationEnterprise Platforms & Tools What We're Looking For 8-12 years of software engineering experience.Strong Python engineering skills with solid software development fundamentals, architecture, testing, version control, packaging, and production-grade delivery experience.Hands-on experience with at least one Agent Development Kit such as LangGraph, CrewAI, Google ADK, OpenAI Agents SDK, AWS Bedrock AgentCore, or Semantic Kernel.Strong expertise in prompt engineering, context engineering, prompt caching, RAG, context graphs, tool calling, MCP, agent memory, and multi-agent orchestration.Experience building evaluation frameworks and measuring agent quality, safety, and reliability.Strong understanding of APIs, secure integrations, CI/CD, observability, AgentOps, and DevSecOps.Ability to rapidly convert business problems into agentic solutions and thrive in client-facing, ambiguous environments.Strong communication and stakeholder management skills.Proven experience mentoring or leading small engineering teams. Preferred Qualifications Experience across multiple ADKs and agent orchestration frameworks.Experience deploying agents on managed enterprise runtimes such as Vertex AI Agent Engine, Bedrock AgentCore, or similar platforms.Domain expertise in Finance Operations, Supply Chain, HR, Claims, or Compliance.Experience implementing Responsible AI and governance frameworks at scale.A track record of building reusable accelerators, agent platforms, or IP adopted across multiple engagements.