Senior Agentic AI Engineer

Elevate Government Solutions•Washington, DC

About The Position

Elevate Government Solutions is a high-growth technology services company focused on driving technological change in government. The Senior Agentic AI Engineer will lead the technical design, prototyping, and deployment of autonomous and semi-autonomous AI workflows for a federal financial-crimes-focused agency. The role supports an agentic foundation initiative to move from productivity tools toward an autonomous, ontology-driven fraud detection and targeting ecosystem, using an AWS GovCloud or Databricks Lakehouse foundation. The engineer will architect multi-agent systems that triage, tag, and analyze Bank Secrecy Act (BSA) data and Suspicious Activity Reports (SARs), with the goal of shortening analyst investigation cycles. The position requires an active Secret security clearance or higher. It can be performed hybrid in Vienna, Virginia, or remotely, and requires availability during East Coast hours. The education requirement is a bachelor's or master's degree in a relevant technical discipline.

Requirements

Active U.S. Government Secret clearance, with ability to obtain and maintain agency suitability/security vetting.
At least 5 years of software engineering experience with deep expertise in Python.
At least 2 years of hands-on experience designing and operationalizing Generative AI and LLM solutions.
Demonstrated experience implementing agentic workflows, including tool/function calling, memory management, or multi-agent coordination with frameworks such as LangGraph, AutoGen, CrewAI, or Bedrock Agents.
Experience in AWS environments, including Bedrock, SageMaker, OpenSearch, Lambda, Step Functions, S3, and IAM.
Experience with structured vector databases and semantic search pipelines over large unstructured narrative collections.
Ability to work East Coast hours.
Ability to work hybrid in Vienna, Virginia, or remotely.
Bachelor’s or Master’s degree in Computer Science, Artificial Intelligence, Data Science, Software Engineering, or a related technical discipline.

Nice To Haves

Direct experience with BSA data, SARs, AML, CTF, or federal fraud analytics.
Hands-on experience integrating knowledge graphs, semantic ontologies, or entity-resolution algorithms such as Senzing, Neo4j, or Amazon Neptune into LLM reasoning loops.
Familiarity with Databricks, Apache Spark, Delta Lake, or lakehouse architectures deployed in federal cloud environments.
Familiarity with Federal AI Executive Orders, the NIST AI Risk Management Framework, and secure enclaves such as FedRAMP High or GovCloud.
Active AWS Certified Solutions Architect or Machine Learning Specialty certification.

Responsibilities

Architect and scale multi-agent systems, including autonomous task routing, iterative planning, reflection, and human-in-the-loop validation.
Develop agents that parse unstructured SAR narratives, extract entities, and apply fraud ontologies for automated tagging.
Implement agent coordination pipelines using frameworks such as LangGraph, AutoGen, CrewAI, or Amazon Bedrock Agents, with deterministic fallback logic and schema enforcement.
Design tools and APIs that connect agent reasoning to the client's Lakehouse platform, BSA datasets, target data selection notebooks, and risk-ranking pipelines.
Integrate graph-based ontologies and entity-resolution services into workflows to trace illicit financial networks, anomalous filers, and typologies across BSA filings.
Support analytic applications including Foreign Target Assessments, Section 311 actions, open request letters, and regulatory comment review.
Deploy agentic solutions on AWS GovCloud and orchestrate foundation models through Amazon Bedrock.
Collaborate with data engineering teams to query and interface with Databricks Lakehouse and Delta tables hosting BSA and SPCL data under active ATO guidelines.
Implement infrastructure as code using Terraform or AWS CloudFormation/CDK, and ensure compliance with federal zero-trust, RBAC/ABAC, and FedRAMP High security baselines.
Design evaluation benchmarks for agent tool-selection accuracy, hallucination rates, and ontology alignment.
Support SI Operations & Maintenance by calibrating guardrails, managing model risk management documentation, and evaluating new foundation model releases.