AI/Data Architect (TS/SCI with Full Scope Poly)

Active TS/SCI with full-scope polygraph required on day one. We can't sponsor or upgrade clearances. This is a hands-on, on-site role in Chantilly, VA.

Who we are

Dimensional Analytics builds and deploys AI for the intelligence community. Small team, senior people, real products in front of real users.

The work

We're taking an AI orchestration platform from R&D to production on a classified network. Agents, tools, and skills reach mission data through a governed gateway, and we need an architect to design how it holds together under accreditation. You'll design it and write a meaningful share of the code yourself. This is not a program management, governance, or advisory role.

We're filling two seats, one weighted toward AI and one toward data. Tell us which one you are.

If you lean AI

• Design the agent and tool architecture: the gateway, MCP servers, skills, and isolated runtimes
• Set model strategy across hosted APIs, Bedrock, and local open-weight models, including when the available API supports only part of what commercial tools expect
• Build the evaluation and approval pipeline that gets new tools and workflows to users safely
• Apply agents to legacy code remediation at scale

If you lean data

• Design how agents reach data across systems with different handling rules: labeling, access policy enforced before every tool call, failing close.
• Define data contracts, lineage, and audit logging that satisfy an ATO
• Tie access to enterprise identity (SSO, LDAP groups) so permissions follow the user
• Set patterns for retrieval, vector search, and structured data access

What we need

• Active TS/SCI with full-scope polygraph
• 8 years building software or data systems, with hands-on LLM or agent work in the last two years
• A system you designed that reached production, and code you wrote for it
• Experience in IC cloud environments (C2S/SC2S) and working within ATO/RMF constraints
• You can explain a design to a government lead and an ISSM in the same meeting
• You use AI coding tools every day

Nice to have

• MCP or agent framework experience
• Kubernetes and GitOps
• Cross-domain or multilevel security background
• You've killed a design you loved because the accreditor said no, and the second version was better

How to apply

Send your resume and a few sentences on one system you designed that went to production: what it did, which parts you built yourself, and what you'd do differently. If it used LLMs or agents, tell us which models and how you served them. We read every one.

Why Dimensional

• New-build engineering on a funded, multi-year program. Not staff augmentation, not O&M.
• You work directly with the founder. No recruiters, no account managers, no layers.
• 8% 401(k) contribution, 22 days of PTO, 100% of your health premium paid, and referral bonuses.
• Benefits: 401(k), 401(k) matching, dental insurance, health insurance, paid time off, parental leave, referral program, retirement plan, vision insurance

Work location: In person, Chantilly, VA

Pay: $245,000.00 - $300,000.00 per year

Benefits:

• 401(k)
• 401(k) matching
• Dental insurance
• Health insurance
• Paid time off
• Parental leave
• Referral program
• Retirement plan
• Vision insurance

Work Location: In person

Salary : $245,000 - $300,000