AI Engineer and Architect

** Job Family:
** Data Science & Analysis
*
* Travel Required:

** Up to 10%
*
* Clearance Required:

** Active Secret

Job Summary The AI Engineer supports system-specific implementations that embed AI/ML and agentic automation to accelerate metadata identification, extraction, enrichment, and documentation. This role owns the metadata harvesting plan and leads the design and implementation of AI-assisted workflows that generate validated metadata artifacts (e.g., semantic tags, data dictionary entries, and supporting documentation) using client-approved tools and human-in-the-loop steward review cycles.

The AI Engineer is accountable for automated metadata harvesting and documentation, delivering automated harvesting and AI-assisted documentation outputs while operating within approved governance and security guardrails.

Key Responsibilities
* Design AI-assisted metadata harvesting & enrichment:
Build AI/ML-enabled approaches to identify, extract, normalize, and enrich technical, business, and operational metadata from structured and semi-structured sources (e.g., databases, pipelines, file/document repositories) and generate semantic tags and documentation outputs.
* Implement agentic/GenAI workflows for metadata documentation:
Design and implement agentic AI patterns to support autonomous or semi-autonomous metadata exploration, summarization, and documentation generation—while maintaining human validation and auditability.
* Own the metadata harvesting plan:
Define scope, sequencing, cadence, source coverage, extraction methods, staging, validation, and handoffs; maintain decision logs and traceability for stewardship and governance review.
* Human-in-the-loop steward validation:
Facilitate structured in-person/virtual review cycles with data stewards to validate AI-generated metadata, resolve discrepancies, and continuously improve extraction/enrichment accuracy.
* Responsible automation scoping:
Identify where automation is feasible vs. where systems require manual curation; document constraints and remediation needs without attempting to “automate through” non-harvestable environments.
* Build and maintain scalable pipelines (metadata-focused):
Implement and maintain scalable pipelines that integrate structured and unstructured sources for metadata extraction and enrichment; apply strong engineering discipline for reliability and repeatability.
* RAG / knowledge integration for semantic discovery:
Lead work that connects harvested metadata to semantic search patterns using retrieval-augmented generation (RAG) concepts and, where applicable, knowledge graph integration to improve discoverability and semantic alignment.
* Metadata quality metrics & validation controls:
Define and implement checks for completeness, accuracy, timeliness, and consistency; flag issues for remediation and support governance escalation where required.
* Technical testing & verification:
Plan and execute tests to verify the metadata extraction/enrichment process works reliably across supported sources; contribute to regression coverage and repeatable validation workflows.
* Mentor and elevate engineering rigor:
Mentor junior team members on agentic AI/GenAI engineering patterns, metadata quality practices, and documentation standards.
*
* What you will need:

*** Must be able to OBTAIN and MAINTAIN a Federal or DoD "PUBLIC TRUST". Candidates with an ACTIVE SECRET CLEARANCE OR PUBLIC TRUST or SUITABILITY are preferred.
* Once onboard with Guidehouse, new hire MUST be able to OBTAIN and MAINTAIN a Federal or DoD "SECRET" security clearance.
* Bachelor’s degree obtained.
* 3-5+ years of experience in AI engineering, applied ML, or data/AI solution development in enterprise environments (senior-level expectations aligned to the referenced agentic AI role).
* Strong proficiency in Python; working proficiency in SQL; familiarity with R is beneficial for statistical validation/profiling.
* Hands-on experience with Generative AI (LLMs and/or similar) and Agentic AI architectures applied to real workflows.
* Experience with RAG pipelines, and familiarity with vector databases and knowledge graphs (for semantic retrieval and metadata-driven…