Manager – Forward Deployed Engineer, Agentic AI

Job Description We are looking for a Forward Deployment Engineer – Agentic AI to design, build, deploy, and continuously improve production-grade AI agents within enterprise client environments. This role combines AI Engineering, software engineering, context engineering, agent evaluation, and client-facing solution delivery. You will work directly with client teams to understand business problems, design agentic solutions, integrate agents with enterprise systems and tools, and ensure that agents are reliable, measurable, secure, and production-ready. The ideal candidate is someone who can move beyond building AI prototypes and can take an agentic solution from business problem → architecture → development → deployment → evaluation → continuous improvement. Key Responsibilities: Agentic AI Engineering - Design and develop LLM-powered agents using Claude Code, OpenAI Codex, or similar AI coding agents that are capable of reasoning, planning, using tools, and executing complex multi-step workflows. - Build autonomous and semi-autonomous agents using frameworks such as LangGraph, LangChain, Google ADK, Semantic Kernel, AutoGen, or equivalent technologies. - Implement agent planning, task decomposition, tool selection, execution, observation, retry, and validation workflows. - Build multi-agent systems where specialized agents collaborate to complete complex business processes. - Integrate agents with enterprise applications, APIs, databases, SaaS platforms, and business workflows. Context Engineering - Design and implement context engineering strategies that provide agents with the right information, instructions, business rules, application state, and tool context at the right time. - Build and integrate Context APIs, RAG, memory/state management, structured context, and dynamic context retrieval where appropriate. - Optimize context to improve agent accuracy, reliability, latency, and cost. - Design mechanisms to prevent irrelevant or excessive context from degrading agent performance. - Apply context engineering principles across real-world enterprise workflows and client environments. Agent Harness Engineering - Design and implement agent harnesses that control state, tool access, permissions, guardrails, retries, recovery, validation, and execution. - Build human-in-the-loop workflows for actions requiring approval or oversight. - Implement secure tool execution and appropriate authentication and authorization controls. - Develop verification mechanisms that allow agents to validate whether an action or task was successfully completed. - Integrate agentic workflows with MCP and other enterprise tool protocols. MCP & Enterprise Integration - Develop and integrate Model Context Protocol (MCP) servers and tools. - Connect agents with enterprise systems including databases, APIs, Git/GitHub, Slack, Jira, Artifactory, CRM platforms, and other business applications. - Build reusable tools that allow agents to take actions, rather than simply generate responses. - Design secure and scalable integration patterns for enterprise AI agents. Agent Evaluation & Quality - Design and implement comprehensive agent evaluation frameworks to measure task completion, accuracy, reliability, tool usage, and business outcomes. - Develop evaluation datasets and automated test suites for agentic workflows. - Implement LLM-as-a-Judge alongside deterministic validation and business-rule-based evaluation. - Evaluate agent planning, tool selection, tool-call accuracy, response quality, hallucination, and task completion. - Establish regression testing to identify changes in agent performance as models, prompts, tools, or context evolve. - Develop monitoring and observability capabilities for production agent behavior. Production AI Engineering - Build production-grade AI services using Python, FastAPI, and modern software engineering practices. - Deploy AI agents within client cloud and enterprise environments. - Build scalable architectures for concurrent users, long-running workflows, and complex agent execution. - Implement caching, asynchronous processing, parallel execution, model optimization, and cost controls. - Integrate AI applications with Git, CI/CD, Docker, cloud platforms, logging, monitoring, tracing, and observability systems. Forward Deployment & Client Engagement - Work directly with client teams to understand business problems, technical environments, data, workflows, and constraints. - Translate business requirements into practical AI and agentic solutions. - Work alongside client engineering, data, product, and business teams to deploy solutions within their existing technology ecosystem. - Troubleshoot production issues and continuously iterate based on client feedback and real-world agent performance. - Communicate technical concepts, trade-offs, limitations, and AI-generated insights effectively to both technical and business stakeholders. - Take ownership of outcomes from initial discovery through production deployment and adoption. Business Impact & Continuous Improvement - Identify opportunities where agentic AI can automate complex workflows, improve decision-making, reduce manual effort, or enhance customer and employee experiences. - Measure the business impact of deployed AI solutions. - Continuously improve agents based on evaluation results, production telemetry, user feedback, and changing business requirements. - Help convert successful AI concepts into repeatable, scalable enterprise solutions.