AI Architect & RAG Data Science Engineer

Modern Technology Solutions Inc•Huntsville, AL

7d•Hybrid

About The Position

MTSI is seeking a "Big Data" Scientist and AI Architect to provide technical leadership on an ongoing and funded MTSI "STRIKE" IRAD. This IRAD focuses on establishing data taxonomy, data meta-tagging, and quality data processes for large, disorganized test datasets, and tailoring AI Agents to quickly find essential data packages for model development. This position will offer key data expertise to MTSI's broad Department of War (DoW) customer base as they tackle these critical challenges. The role serves as an AI Architect and RAG Data Science Engineer, supporting the design, delivery, and continuous improvement of a secure Retrieval-Augmented Generation (RAG) capability. The individual will be responsible for designing secure, full-stack AI/ML architectures that encompass data ingestion, APIs, model services, retrieval, applications, and deployment infrastructure. The role translates mission needs into governed data products and full-stack AI services, converting distributed technical, engineering, test, and telemetry information into traceable, role-appropriate answers and analytics. Collaborating with technical and program leadership, the incumbent will contribute hands-on expertise across AI/ML, software, cloud, cyber, data engineering, systems engineering, and testing. The mission focus is to deliver trustworthy, secure, and measurable AI-enabled knowledge access and decision support across constrained enterprise and mission environments.

Requirements

Four or more years of professional experience delivering software, data, analytics, AI/ML, cloud, or data-platform capabilities in an enterprise, regulated, mission-critical, or defense-adjacent environment.
Bachelor's degree in artificial intelligence, machine learning, data science, computer science, engineering, applied mathematics, or a related discipline; equivalent relevant experience may be considered where permitted by contract.
Deep practical understanding of agentic AI and RAG systems, including tool use, multi-step orchestration, retrieval design, prompt and model orchestration, source traceability, evaluation, and responsible-AI controls; experience with llama.cpp and OpenCode or comparable approved tools.
Working knowledge of search and retrieval algorithms including BM25, TF-IDF, vector embeddings, hybrid retrieval, metadata filtering, and reranking.
Strong programming and engineering foundation using Python and at least one additional object-oriented language such as Java, C#, or comparable technologies; familiarity with Git, debugging, build workflows, and code review practices.
Working knowledge of SQL, structured and unstructured data pipelines, APIs, service-oriented architectures, document processing, vector search, web or frontend integration, containers, CI/CD, and secure software development practices.
Ability to communicate technical designs, delivery risk, test evidence, and operational implications clearly to technical, government, contractor, and nontechnical stakeholders.

Nice To Haves

Master's degree or active graduate-level study in AI, ML, data science, computer science, engineering, or a closely related field.
Experience with Kubernetes, cloud-native platforms, infrastructure as code, GitLab, GitHub Actions, Azure DevOps, or comparable automated delivery toolchains.
Experience operating AI, analytics, telemetry, digital engineering, test, or sustainment solutions in Department of Defense, aerospace, aviation, or similarly constrained environments.
Experience implementing model evaluation, observability, safety or policy guardrails, identity-aware access controls, and data or model provenance for generative-AI systems.
Experience leading Agile or hybrid delivery teams and integrating AI products with enterprise systems, secure networks, edge deployments, or disconnected operations.

Responsibilities

Translate operational, test, and sustainment needs into RAG solution designs, delivery increments, acceptance criteria, and measurable decision-support outcomes under established program priorities.
Design and implement secure RAG and agent capabilities, including source onboarding, document parsing and normalization, metadata and taxonomy management, hybrid retrieval, reranking, LLM inference, citations, guardrails, and agentic workflows using controlled tool use and handoffs; apply llama.cpp or comparable local inference runtimes where permitted.
Apply search methods such as BM25, TF-IDF, embeddings, hybrid retrieval, metadata filtering, and reranking to enhance recall, precision, traceability, and user trust in technical knowledge retrieval.
Develop evaluations for data quality, retrieval relevance, groundedness, faithfulness, latency, usefulness, and safety; maintain curated test sets, analytic baselines, experiments, and performance dashboards to support evidence-based releases.
Contribute to secure user experiences, RESTful APIs, application services, workflow orchestration, data stores, vector databases, integration patterns, and observability required for operating an AI product at enterprise scale.
Integrate structured and unstructured technical data, test artifacts, logs, sensor or platform telemetry, and operational knowledge while preserving provenance and access controls, in partnership with engineering, test, and data owners.
Apply DevSecOps and MLOps principles, including Git-based development, automated testing, CI/CD, containerization, vulnerability management, model and data versioning, monitoring, auditability, and repeatable deployment across approved environments.
Decompose ambiguous technical problems involving fragmented data, conflicting sources, constrained networks, evolving requirements, performance tradeoffs, and mission risk; document findings and communicate workable alternatives to technical and program stakeholders.