Senior AI Engineer – Forward Deployed, Agent Platform (REMOTE)

The Home Depot•Atlanta, GA

5d•$100,000 - $180,000•Remote

About The Position

The Senior AI Engineer – Forward Deployed, Agent Platform embeds with The Home Depot’s business verticals—including Finance, Marketing, Supply Chain, Online, Store Operations, HR, and Legal/Risk—to turn ambiguous operational workflows into production-ready, enterprise-governed AI agents. The role consults with business partners on technical approaches, teaches them how to build and manage agents across available platforms, and promotes enterprise adoption and reuse of agents, Skills, and tools through a shared marketplace. The position builds and deploys multi-step agents and conversational assistants, enterprise integrations and MCP servers, reusable evaluation harnesses, and templates and standards that make subsequent builds more efficient. It reviews agent submissions for engineering, security, and Responsible AI quality; supports ownership handoffs; enables business and product teams; and turns recurring needs into scoped platform backlog items. Foundation models are consumed as managed services: pre-training or fine-tuning models, GPU/CUDA kernel work, and classical feature stores or model registries are outside the role’s scope. The role’s work is described as 70% delivery and execution, 20% support and enablement, and 10% learning. It typically reports to a Software Engineer Manager or Sr. Software Engineer Manager and has no direct reports. Overnight travel is typically required 5% to 20% of the time. The position is remote. The work is primarily in a comfortable indoor setting, with most time spent sitting and occasional movement; rarely, light articles may need to be moved or lifted.

Requirements

At least 18 years of age.
Legally permitted to work in the United States.
Minimum of 2 years of work experience.
High school diploma and/or GED, as the minimum education typically acquired for the role.
Preferred experience includes 6 or more years shipping and operating production Python services, including FastAPI, asynchronous microservices, and REST.
Experience with multi-step agent orchestration (such as Google ADK, LangGraph, CrewAI, AutoGen, or equivalent) and LLM APIs such as Gemini, Claude, or OpenAI.
Systematic prompt engineering, structured outputs, context optimization, and evaluation experience.
Enterprise SaaS and MCP integration experience, including OAuth2, SSO, service accounts, delegated-user access, and RBAC boundaries.
Experience with RAG, knowledge ingestion, and search, including access-scoped retrieval.
Experience building agent evaluation harnesses and golden datasets, and applying high-availability and defensive-coding practices.
Experience with GCP, including Vertex AI, Cloud Run, and BigQuery; analytical SQL; CI/CD; and observability tools such as Grafana, Prometheus, or OpenTelemetry.
Forward-deployed execution skills, including cross-functional collaboration, code and prompt review, enablement documentation, and handoffs.
Demonstrated experience working with nontechnical stakeholders, comfort with ambiguity, and ability to present technical tradeoffs to executives; customer-facing communication is assessed through an interview mock discovery session.
Security and Responsible AI fundamentals for agents that take write actions, including PII handling, human-in-the-loop approvals, audit trails, and least-privilege access.

Nice To Haves

Hands-on experience with Gemini Enterprise, Vertex AI Search or Data Stores, Google ADK, and MCP or FastMCP.
Prior experience as a Forward Deployed Engineer, Solutions Engineer, or Embedded Platform Engineer.
Experience with human-in-the-loop approval workflows, write-action safety controls, and immutable audit trails.
Experience with production Responsible AI and safety guardrails, including Model Armor, prompt-injection defense, bias and hallucination mitigation, and PII redaction.
Experience with A/B testing, graph or connected knowledge systems, Terraform/HCL, Docker/GKE, and TypeScript/React/JavaScript for UI or Canvas surfaces.
Retail, eCommerce, supply chain, store operations, or finance domain background.

Responsibilities

Partner with non-technical business stakeholders to discover workflows, define requirements and negative scope, and establish acceptance criteria before development.
Build and deploy production multi-step agents and conversational assistants using Python, FastAPI, Google ADK, and Gemini Enterprise, including human-in-the-loop confirmation gates for write actions.
Build resilient MCP servers and REST integrations for enterprise systems, with OAuth/IAM boundaries, idempotency, retries, and high-availability patterns; publish reusable integrations to the enterprise marketplace.
Create workflow-level golden datasets, agent and prompt evaluations, and automated regression suites that provide business owners with objective pass/fail criteria.
Extract reusable templates, Skills, and packaging standards from projects to reduce the effort required for subsequent builds.
Review partner-submitted agent code, Skills, and prompts for engineering, security, and Responsible AI standards; curate the marketplace so published assets are discoverable, documented, and trusted.
Write runbooks and enablement materials, verify the receiving owner is ready, and transfer day-to-day operations to the relevant business vertical.
Translate recurring vertical needs into scoped backlog items for the APEX platform, including Gateway/Registry, Model Armor guardrails, observability, and lifecycle management.
Support technical workshops and office hours, review low-code agents, maintain documentation and a pattern library, and use usage data and feedback to improve adoption and the platform.
Collaborate with UX, engineering, and product management to create secure, reliable, and scalable machine-learning solutions; ensure quality and change-control standards are met.
Work with product teams to make user stories developer-ready, understandable, and testable.
Write code or scripts to automate infrastructure, monitoring, test cases, and destructive resiliency testing.
Configure commercial off-the-shelf solutions to meet changing business needs; create dashboards, logging, alerting, and response mechanisms to proactively capture and address issues.
Participate in learning activities and communities of practice; seek out material on modern software design, machine learning, development practices, and emerging technologies.
Answer questions from product and support teams, encourage collaboration, provide production application support, monitor service-level objectives, and review production performance and capacity across code, infrastructure, data, message processing, and prediction quality.