> Markdown version of [/jobs/ext/3869026-architect-ai-active-top-secret-clearance](https://www.wearedevelopers.com/jobs/ext/3869026-architect-ai-active-top-secret-clearance). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Architect, AI - Active Top Secret Clearance - **Company:** Red Hat - **Location:** United States (Remote available) - **Experience:** Expert - **Salary:** $130,420.0 - $215,300.0 - **Contract:** Permanent contract - **Skills:** LangGraph Framework, Java (Programming Language), Application Programming Interfaces (APIs), Artificial Intelligence, Continuous Integration, Python (Programming Language), Open Source Technology, OpenShift, Role-Based Access Control, Red Hat Enterprise Linux, OpenAI, Ansible, Software Engineering, Retrieval-Augmented Generation, Large Language Models, Agentic-AI, Kubernetes, Machine Learning Operations, VLLM, KServe, Data Generation, Agent2Agent Protocol - **Published:** October 22, 2026 - **Apply:** https://startup.jobs/architect-ai-active-top-secret-clearance-red-hat-srl-10544426 ## About the Role The Red Hat Services team is looking for an AI Architect to design and deliver strategic solutions that solve our clients' business challenges, through technical leadership, hands-on implementation, and enablement of recommended practices. In this role, you will play a key role as a promoter of the adoption of Red Hat's AI platforms and solutions by using your technical expertise and interpersonal skills. You'll need to have a balance of solid technical experience, team leadership, awareness of open source, thorough understanding of business processes, and the ability to identify and solve issues at an enterprise level., * 10+ years in enterprise technology, with at least 5 as a hands-on architect or senior consultant delivering in customer environments. * Deep, current OpenShift expertise: cluster architecture, GitOps, multi-cluster management, security and observability, GPU-enabled nodes; Red Hat Certified Architect or equivalent demonstrated depth. * Production experience with Red Hat AI or its upstreams: OpenShift AI, vLLM, model serving with KServe, model registries and pipelines, and at least one of llm-d, LLM Compressor or Eval Hub. * Working knowledge of how enterprises run generative AI today: inference economics, model selection and optimization, RAG and evaluation, agent frameworks and harnesses, MCP, and the governance and cost pressures that shape production decisions. * Application development fluency on Kubernetes: containers, APIs, CI/CD and the trusted software supply chain (SBOM, signing, policy gates), in at least one of Python, Go or Java. * Automation with Ansible Automation Platform, including event-driven automation, or an equivalent you can map onto it quickly. * A record of leading engagements to a measurable result on a date, and of leaving customer teams able to run what you built. * Clear writing and speaking for technical and executive audiences; comfortable setting a success metric with a sponsor and reporting against it. ## Description This position requires candidates to work on-site with clients, so a willingness to travel regularly to customer locations throughout North America is required. Successful applicants must reside in a state where Red Hat is registered to do business and hold an Active Top Secret Clearance (TS/SCI CI Poly). What you will do Our customers want to make AI their own: their model, their data, their AI. Help customers make frontier and open models their own * Lead model selection, optimization and serving decisions on Red Hat AI: vLLM, llm-d, LLM Compressor, the validated model catalog, and hardware profiles across NVIDIA, AMD and other accelerators. * Baseline and reduce inference cost per use case (cost per million tokens, GPU utilization) using GuideLLM benchmarks and recipe-based optimization, and decide with evidence whether to prompt, ground or tune a model for a task. * Design portable inference: workloads that move between cloud APIs, on-premises clusters and sovereign environments without re-architecture. * Travel weekly, Monday-Thursday, to client locations Ground and evaluate on the customer's terms, in the customer's environment * Design evaluation-driven release gates in Eval Hub so a model or agent ships only when the customer's own tests say so. * Architect retrieval and context patterns (RAG, knowledge ingestion, permission-aware context) that keep data where it belongs, and know when that work is repeatable and when it is bespoke. * Lead sovereign and disconnected deployments of Red Hat AI on OpenShift, including model provenance, signing, AI bills of materials and compliance evidence. Run it as a service, and run operations with it * Design and stand up Models-as-a-Service and GPU-as-a-Service on OpenShift AI and Red Hat AI Enterprise: multi-tenancy, the AI gateway, metering, showback and chargeback, self-service accelerator access. * Build agentic systems that operate infrastructure through Ansible Automation Platform: patching, CVE remediation, upgrades, supply-chain attestation and incident remediation, with every action passing through RBAC, approval and audit. * Apply agent frameworks and protocols as customers use them today (MCP, A2A, OpenAI-compatible APIs, Llama Stack, LangGraph and similar), with identity, policy and observability designed in from the start. Lead the engagement and leave it better * Own the architecture and the delivery plan on engagements from two-week baselines to multi-month service builds and lighthouse programs; set the success metric with the customer's sponsor and hold the team to it. * Pair with and certify customer engineers so the result keeps running after we leave. * Turn each delivery into a reusable asset for the practice: runbooks, reference architectures, harnesses and recipes, and feed product gaps back to Red Hat engineering with evidence. * Represent Red Hat's point of view on enterprise AI with CTOs, platform leaders and FinOps teams, and support sellers in scoping results that can be dated with confidence. * Provide customer feedback to internal teams to enhance product offerings and services. * Lead internal initiatives that foster innovation, efficiency, and growth within the Services business. * Continuously learn, grow, and adapt to new skills and technologies * Regular travel to client locations, up to weekly according to project needs, * Agent development in production: tool calling, agent identity and delegation, sandboxed execution, agent observability and evaluation; contributions to MCP, A2A, Llama Stack, LangGraph, Kagenti or similar. * Experience designing Models-as-a-Service or GPU-as-a-Service for shared platform teams: multi-tenancy, gateways, metering, showback and chargeback, scheduling with Kueue or Run:ai. * FinOps for AI: building cost baselines, unit economics per use case, and verifying savings on a customer's own hardware. * Sovereign, regulated or disconnected deployments; familiarity with EU AI Act obligations and model provenance requirements. * Model customization: fine-tuning and post-training, synthetic data generation, and the judgment to know when prompting or grounding is the better answer. * Open source contribution and community standing in projects Red Hat leads or sponsors. * Prior forward-deployed, residency or customer-engineering work at a platform or model company.