How to apply for AI Solutions Architect- Federal

About HiddenLayer

HiddenLayer builds security for machine learning models and algorithms. Its platform covers threat modeling, ML risk assessment, red teaming, and AI/ML model scanning. The company was founded by AI and security practitioners and has won RSA's Innovation Sandbox, CB Insights AI 100, and CyberTech 100. You would work on defending AI workloads, a problem the founders saw first-hand.

About the role

This is the pre-sales technical lead for federal accounts across the Intelligence Community, DoW/DoD, DHS, and civilian agencies. You partner with Account Directors to move opportunities from first discovery through proof of value and technical win. You architect and deploy the platform in SaaS and airgapped environments, write integrations against the SDKs and APIs, and build demonstrations that show how AI workloads are attacked and defended.

A typical day

A typical day mixes customer-facing discovery or demo calls with hands-on work building integrations, demos, or proof-of-concept deployments of the HiddenLayer platform. You coordinate with Account Directors on pursuit strategy and with internal teams on technical requirements for airgapped federal environments. Exact routines and team structure are not stated in the job details, so ask about them in the interview.

Who HiddenLayer is looking for

• Holds an active Top Secret clearance or is TS eligible, with DC Metro area proximity preferred.
• Has pre-sales or solutions architecture experience selling technical products into federal agencies, including the IC, DoW/DoD, DHS, or civilian agencies.
• Can write real code against SDKs and APIs, not just describe architecture, and can build proof-of-concept AI applications.
• Understands AI/ML workloads and adversarial AI attacks, and can explain them to mission owners and technical evaluators.
• Has deployed or supported systems in disconnected, airgapped environments.

Tips for this application

• State your clearance status clearly and early: active Top Secret, TS eligible, or neither. This is a screening criterion.
• Describe at least one federal deal you carried from discovery to technical win, including the agency type and the technical objection you overcame.
• Show code or integration work. Link a repo, describe an SDK integration you wrote, or explain a proof-of-concept you built against an API.
• Address airgapped or disconnected deployment experience directly, since the role requires it alongside connected SaaS.
• Reference HiddenLayer's specific products (model scanning, red team assessment, ML risk assessment) and connect them to a federal mission problem you have seen.

What to cover in your cover letter

['Your clearance status and eligibility, plus any DC Metro area ties.', 'A concrete example of serving as pre-sales technical lead on a federal pursuit, from first discovery call to proof of value.', 'Hands-on work with SDKs or APIs, especially building demos or proof-of-concept AI applications.', 'Experience deploying or operating systems in airgapped or fully disconnected environments.']

Research before applying

• Read HiddenLayer's product pages on model scanning, ML risk assessment, and red team assessment to understand what you would demo.
• Review the RSA Innovation Sandbox, CB Insights AI 100, and CyberTech 100 recognition and the specific work that earned them.
• Study recent federal AI policy and guidance for the IC, DoW/DoD, DHS, and civilian agencies to speak to mission owner priorities.
• Look into common adversarial AI attack types (for example evasion, poisoning, model extraction) so you can discuss them concretely.

Likely interview topics

Based on the job description, expect questions about:

• How you would run a discovery conversation with an Intelligence Community or DoD mission owner who is new to AI security.
• How you would design a proof of value for the HiddenLayer platform in an airgapped environment.
• A walkthrough of an integration you wrote against an SDK or API, including problems you hit.
• How you explain adversarial AI attacks and model risk to both technical and non-technical federal stakeholders.
• How you would handle a technical objection or a competing solution during a federal evaluation.

Common mistakes to avoid

• Being vague about clearance status. Federal pre-sales screens on this first.
• Presenting only slide-deck pre-sales experience without evidence of writing code against SDKs or APIs.
• Ignoring the airgapped requirement and assuming all deployments are standard cloud SaaS.
• Treating AI security as generic cybersecurity. This role requires understanding attacks on ML models specifically.