Technology Strategy Director

Freddie Mac•McLean, VA

About The Position

Freddie Mac is seeking a hands-on AI engineering leader to serve as Director within the Single-Family Acquisitions (SFA) division. Reporting to the Sr. Director of GenAI Engineering & Enterprise Rules Platform, this player/coach role owns the technical core of SFA’s AI Systems Solutioning, including agent engineering and design. The work spans the design and execution of AI systems from ideation and business casing through specification, implementation, and deployment; evaluation harnesses and graded assessments that determine production readiness; and the operating layer that keeps agents accountable after launch. The role is approximately 70% hands-on architecture and code and 30% leadership of GenAI engineering efforts and platform and vendor delivery partners. It is forward-deployed by design, partnering with SFA business teams, the divisional GenAI PMO, Enterprise Operations & Technology (EO&T), Digital Experience, and the Enterprise Business Rules platform team. The Director will establish the engineering discipline and shared capabilities needed to move agentic workflows from pilot to production and operate them reliably. This includes durable orchestration, measurable accuracy, operational transparency, retrieval, evaluation, monitoring, and knowledge graphs representing business functions. The role will develop shared engineering patterns and a common domain vocabulary, evaluation approach, and traceability model for agent-driven and rule-driven decisioning. The position leads GenAI engineering delivery while remaining a senior hands-on engineer. It establishes a forward-deployed engagement model, institutionalizes spec-driven development, builds reusable patterns and reference implementations, mentors engineers, and communicates with business owners and senior leadership. The Director will architect and deliver an agentic workflow platform, shared domain graph, evaluation harnesses, governed hybrid retrieval, and post-launch operational capabilities. The role will personally deliver initial production use cases in credit risk and quality control from architecture through implementation, operations, and iteration. The role embeds Responsible AI engineering, including content guardrails, adversarial red-teaming, bias and fairness reviews, model validation, explainability, and governance artifacts for regulated release gates. Agent decisions must be traceable to source documents, prompt versions, and evaluation evidence, and solutions must comply with model risk management, data protection, and regulatory requirements applicable to GSE decisioning. Success in the first 12–18 months includes putting the agentic workflow platform into production with at least two SFA use cases, adopting evaluation harnesses and release gates as the standard for agent promotions, launching observability with trace lineage, cost telemetry, SLO dashboards and runbooks, putting the shared domain graph in use across agentic workflows and the Business Rules platform, and measurably reducing time to production for new agentic capabilities.

Requirements

12+ years building and operating production ML and AI systems in regulated environments.
5+ years leading engineering teams.
Demonstrated hands-on delivery of a multi-agent orchestration platform to production using LangGraph or equivalent, including durable workflow state and declarative agent configuration.
Experience building evaluation harnesses, graded datasets, and release gates for LLM systems and operating them in LangSmith or equivalent tracing and observability tooling.
Production-scale knowledge graph and hybrid RAG experience with Neo4j, vector search, and governed document corpora.
Track record of spec-driven development, writing specifications and acceptance criteria first and building with AI coding assistants against them.
Production Python experience.
College degree or equivalent experience.
Comfort embedding with business teams and remaining hands-on in code.
Ability to demonstrate the value of delivered AI systems and to treat specifications and evaluation sets as first-class deliverables.
Systems-thinking approach that designs for future use cases as well as current needs.
Clear communication with engineers, business owners, and executives.

Nice To Haves

TypeScript/React experience.
AWS (Bedrock) or Azure experience.
Advanced studies preferred.
Financial services or mortgage domain experience strongly preferred.
Experience with LangGraph Platform, AutoGen or equivalent, MCP, agent-to-agent handoffs, AWS Bedrock, OpenAI, Anthropic, Graph RAG, OpenSearch or equivalent, LangSmith, and GitHub Copilot, Amazon Kiro, or OpenCode.

Responsibilities

Lead GenAI engineering efforts, execution, and delivery partners under a single delivery model while remaining the senior hands-on engineer.
Establish a forward-deployed engagement model that embeds AI engineers alongside SFA business teams and remains accountable for production outcomes and post-launch behavior.
Institutionalize spec-driven development, ensuring specifications, acceptance criteria, and evaluation targets are written before development begins.
Establish and maintain a pattern library of reusable kickstarters, skills, and reference implementations, using AI coding assistants within the spec-driven workflow.
Partner with the divisional GenAI PMO on capability roadmaps, prioritization, and enablement; partner with EO&T on reference architectures and platform standards; and partner with Digital on product ownership and delivery practices.
Mentor engineers on agent design, evaluation discipline, and production operations, and serve as a technical voice with senior leadership on agentic readiness.
Architect and deliver a durable agentic workflow platform with persistent state, declarative agent configuration, tool and API integration, and agent interoperability.
Design a shared domain graph and knowledge and service layer as a common vocabulary across agents and rules, enabling validation across workstreams without coupling them.
Build assessment harnesses and graded datasets to gate promotion, including scored runs, full trace capture, and regression checks; use the harnesses for prompt iteration and model selection.
Establish hybrid retrieval over a governed document substrate, with docs-as-code ingestion and approvals, reusable ingestion, guardrails, and observability.
Engineer the post-launch operating layer, including trace analysis, token and cost telemetry, SLO dashboards, drift detection, runbooks, monitoring, and documentation for business owners.
Personally deliver initial production use cases in credit risk and quality control from architecture through implementation, operations, and iteration.
Embed Responsible AI engineering across solutions, including content guardrails, adversarial red-teaming, bias and fairness reviews, model validation, and governance artifacts.
Ensure agent decisions are traceable to source documents, prompt versions, and evaluation evidence.
Ensure compliance with model risk management, data protection, and regulatory requirements applicable to GSE decisioning.