TixelJobs
S
Strategic ACIvia muse

Senior AI Platform Engineer

Springfield, VA, USPosted 2w ago
ML EngineerSeniorFull-time

About the Role

Job Description



Strategic ACI is seeking a Senior AI Platform Engineer to design, build, deploy, and operate production AI capabilities for NGA on a new Digital Ecosystem program. As a member of the Platform/Cloud team, this individual will develop the reusable services, integrations, agent harnesses, and operational practices that enable large language models (LLMs) and AI agents across the customer’s broader Digital Ecosystem. You will help develop an enterprise architecture digital twin, connecting UAF and SysML models with real\-time and near\-real\-time operational data. These capabilities will enable AI\-assisted analysis, data\-driven insights, and visualizations of enterprise capabilities, dependencies, risks, and mission impacts.

The role combines architecture with direct implementation, including writing code, integrating models and tools, building deployment pipelines, evaluating AI performance, and troubleshooting production systems. Working alongside platform engineering, cybersecurity, enterprise architecture, and mission teams, this individual will turn emerging AI technologies into secure, reliable capabilities that support customer workflows.

Salary will be commensurate within the advertised range, based on applicable qualifications.



**Responsibilities:**

**AI Platform Development:** Design and implement reusable AI services, APIs, components, and reference architectures that support multiple applications and mission workflows across the Digital Ecosystem.

**LLM Integration and Engineering:** Build LLM\-enabled capabilities using model APIs, retrieval\-augmented generation, prompt and context engineering, structured outputs, and inference services. Evaluate hosted, self\-hosted, and open\-weight models against security, performance, deployment, and cost requirements.

**Agent Harnesses and Workflows:** Develop agent harnesses that manage model interactions, context, state, memory, tool access, workflow execution, and error handling. Implement bounded autonomy, human approval, execution limits, and recovery mechanisms for agents interacting with enterprise systems.

**Production Deployment and Operations:** Build automated testing and deployment pipelines for AI services and agent workflows. Implement versioning, configuration management, monitoring, rollback, incident troubleshooting, and continuous improvement. Address reliability, latency, resource consumption, and operational cost.

**AI Evaluation and Observability:** Develop representative evaluation datasets and automated tests for accuracy, groundedness, task completion, safety, and tool\-use behavior. Instrument AI services to trace model interactions and agent actions, support troubleshooting, and detect performance degradation.

**AI Governance and Cybersecurity:** Implement controls for approved models and tools, identity, least\-privilege access, secrets management, sensitive data handling, audit logging, and third\-party dependencies. Address prompt injection, data leakage, insecure tool execution, and AI supply chain risks in coordination with cybersecurity and governance teams.

**Enterprise Architecture Digital Twin:** Work with enterprise architects and digital engineering teams to build a digital twin that connects UAF and SysML models with enterprise applications, infrastructure, services, and operational information. Support an evolving view of enterprise capabilities, dependencies, and conditions.

**Real\-Time Model Integration:** Build interfaces and event\-driven integrations that associate real\-time and near\-real\-time feeds with UAF and SysML model elements. Apply AI to correlate information, identify changes and inconsistencies, and propose validated model updates while preserving provenance, configuration control, and human oversight.

**AI\-Enabled Insights and Visualizations:** Combine architecture models and operational data to identify patterns, anomalies, dependencies, risks, and potent
Share