AI/ML Data Engineer

Precise Software Solutions Inc. is seeking an AI/ML Data Engineer to design and govern the data foundation behind secure AI-enabled applications for the U.S. Government. You will own AI-ready datasets, from ingestion and quality controls through privacy protection, lineage, and retrieval-quality measurement, so that every AI output can be traced back to trusted source data.

Location: Washington, DC metro area (hybrid) 3 days a week on site in DC

Citizenship: U.S. citizen or green card holder/EAD

Background Investigation: Eligible for public trust clearance- Government background investigation required

Key Responsibilities:

• Design and maintain data ingestion, transformation, and processing pipelines (ETL/ELT) for AI training, evaluation, retrieval, and operations, including support for data migration and cleansing.
• Curate, validate, and version datasets, and maintain dataset inventories, metadata, lineage, provenance, and ingestion logs.
• Implement automated data-quality checks for duplication, schema changes, completeness, and freshness, and maintain dataset quality scorecards and drift reports.
• Design data models, vector stores, and embedding schemas for Retrieval-Augmented Generation (RAG) knowledge bases, and re-index content when sources change.
• Measure retrieval and model quality against established baselines using metrics such as precision/recall, MRR, NDCG, and context relevance.
• Prepare data-related deliverables, including AI model cards, ML and AI pipeline documentation, RAG/AI Pipeline Evaluation Reports, data dictionaries and embedding schema documentation, and responses to Government data calls.
• Build secure structured-data access for AI applications (e.g., natural-language-to-SQL with query validation and role-based authorization), and support dashboards and operational analytics.
• Monitor data and ML pipelines, troubleshoot failures, and support root-cause analysis, while keeping all data in FedRAMP-authorized cloud regions with FIPS-validated encryption.
• Support Responsible AI practices by preparing representative evaluation datasets, testing AI outputs for bias, accuracy, and hallucination, and documenting results to meet federal AI governance requirements.
• Secure the AI data path, from source datasets and embeddings to prompts and logs, and support AI risk testing such as data poisoning.