Government and Infrastructure - AI and Data - AI Automation Engineer - Manager

EY•McLean, VA

1d•$125,500 - $261,600•Hybrid

About The Position

The AI & Automation Engineer designs, builds and operates the AI and automation capabilities that allow the team to assess a large legacy application portfolio at a speed and scale that manual analysis cannot achieve. This role converts raw source code, database schemas, configuration and technical documentation into structured, evidence-backed portfolio intelligence — component inventories, dependency graphs, capability maps, technical-debt and risk indicators, and consolidation candidates — that directly drive rationalization and modernization decisions. AI & Automation Engineers work side by side with Application Modernization Engineers, who validate automated findings and feed confirmed patterns back into the pipeline, continuously improving accuracy. The capabilities built here become reusable assets for later modernization, enhancement and operations work across the program.

Requirements

Must have a Bachelor's degree in computer science, software engineering, information systems, computer engineering or a related field, or equivalent practical experience.
Must be eligible to obtain and maintain a Secret Clearance
6–10+ years of software, data or ML engineering experience, including architecting and delivering production AI/ML systems, leading teams of five or more engineers, and client-facing delivery in federal or regulated environments
Must be comfortable working in person as needed in the Washington, DC area

Nice To Haves

Understanding of generative AI models (openAI, Anthropic, open-source LLMs) into web-based products
Experience working with designers, product managers, AI/ML engineers to create intelligent user experience
Experience writing tests, participating in code reviews
Seeking out information to learn about emerging methodologies and technologies
Clarifying problems by driving to understand the true issue
Looking for opportunities for improving methods and outcomes
Collaborating, influencing, and building consensus through constructive relationships and effective listening
Solving problems by incorporating data into decision making
Static code analysis and parsing (Tree-sitter, Roslyn, JavaParser or abstract syntax tree tooling).
Graph databases (Azure Cosmos DB Gremlin API, Neo4j) and graph algorithms such as clustering and community detection.
Agent frameworks (LangGraph, LangChain, Semantic Kernel) and Model Context Protocol (MCP).
LLM evaluation and observability tooling.
Azure services including Container Apps, AI Search, Cosmos DB, PostgreSQL, Service Bus, Key Vault, Entra ID and private networking; Azure Government experience.
Experience with legacy application modernization or application portfolio assessment.
Microsoft Certified: Azure AI Engineer Associate (AI-102) or comparable certification.
Prior federal government delivery experience.

Responsibilities

Build and operate automated ingestion pipelines for source code repositories, database scripts and schemas, build and deployment configuration, and technical documentation across heterogeneous technology stacks (.NET, Java, ColdFusion, JavaScript, batch and legacy runtimes).
Develop static-analysis and LLM-assisted extraction to identify application components, technology fingerprints (languages, frameworks, versions and libraries), dependencies, interfaces, data-access patterns and shared database usage.
Build LLM-based capabilities for code and business-logic summarization, business-capability tagging, functional similarity and duplicate detection to surface consolidation candidates, and technical-debt, security and risk indicators.
Implement retrieval-augmented generation (RAG) and hybrid (keyword and vector) search over portfolio evidence so analysts and client stakeholders can query the portfolio in natural language with source citations.
Load extracted entities and relationships into graph and relational data stores to support dependency analysis, change-impact analysis, cluster detection and modernization wave sequencing.
Build evaluation harnesses — golden datasets, grounding and faithfulness checks, accuracy measures and regression tests for prompts and agents — and track quality metrics over time.
Design human-in-the-loop review points and capture reviewer decisions, model and prompt versions, and rationale to maintain end-to-end traceability of AI-supported findings.
Instrument pipelines for observability, throughput and token/cost consumption, and optimize for performance and cost at portfolio scale.
Package prompts, agents, tools and pipeline components as reusable, documented assets for subsequent modernization, enhancement and operations work.
Produce technical documentation, runbooks and knowledge-transfer materials for the capabilities you build.
Lead the AI & Automation team: plan and prioritize work, manage delivery against 90-day milestones and remove blockers.
Own the architecture of the automated assessment platform, including model and tool selection, data flows, security boundary, scalability and cost.
Establish responsible-AI governance for the program — model risk management, transparency, auditability and human-review policy — aligned with federal AI guidance and the NIST AI Risk Management Framework.
Serve as the primary technical point of contact with the client on AI methods; present results and explain methodology to client leadership.
Coordinate closely with the Application Modernization lead so automated findings and manual validation form a single, consistent evidence base.
Recruit, mentor and develop engineers, and contribute to the staffing plan as the program scales.