Product strategy & discovery
Clarify the problem, define the user need, map the workflow, and turn ambiguous requirements into a focused product direction.
Federal Digital Services & AI Product Design
Feedback Loupe designs for consequential systems—public services people have to understand, and mission systems operators have to trust. Senior product support to clarify ambiguous opportunities, validate user needs, and design AI-enabled digital services that are usable, testable, and implementation-ready.
Senior product design and prototyping support for teams working through complex services, AI-enabled workflows, digital modernization efforts, and public-facing user experiences.
Clarify the problem, define the user need, map the workflow, and turn ambiguous requirements into a focused product direction.
Plan and conduct user interviews, synthesize findings, identify decision points, and translate research into product requirements.
Create journey maps, service blueprints, intake flows, handoff models, and operational views for complex service environments.
Design human-in-the-loop workflows, agent experiences, decision-support interfaces, review queues, and trust-building AI interactions.
Create high-fidelity prototypes that help stakeholders test assumptions, validate direction, and de-risk implementation before full build.
Deliver clear interface specifications, design systems, usability improvements, and engineering-ready product experiences.
One craft—stakes, not sector. Lane A carries the proof-backed depth; Lane B extends the same human-in-the-loop design practice into mission and operational systems.
Lane A — Federal Health & Public Services
The stake is comprehension. A service the public can’t understand hasn’t been delivered.
Public health, benefits, eligibility, patient and provider experience, science and health communications, and public-facing digital services—where complex policy or clinical information has to become something a person can read, trust, and act on.
Backed by the principal’s career, not the entity’s. Chris’s prior work—NHTSA’s safety web under Section 508 (via UICGS/Bowhead), NIH-adjacent health communications (via Palladian Partners), and Amtrak’s 30M-user booking system—is the demonstrated depth behind this lane. Feedback Loupe is the practice he now brings it through.
Lane B — Defense, Mission, Cyber & Federal IT
The stake is trust under load. An operator who doesn’t trust the tool routes around it.
Mission software, operational dashboards, analyst and caseworker workflows, cyber workflows, AI-assisted decision support, and human-machine teaming—workflow-heavy systems where clarity, provenance, and operator trust decide whether the tool gets used.
Credibility rests on the principal’s prior environments—defense-IT work via UICGS/Bowhead, plus product design at a national-security venture studio and AI/cyber incubation—not on entity past performance. No active clearance; clean record; CUI-aware. Positioned at the unclassified/CUI design layer, subcontracting under cleared primes—not as a cleared prime.
Evidence of the principal’s capability—not corporate past performance for Feedback Loupe LLC.
Engagements performed by Chris Mullins as an employee or subcontractor of other firms, prior to founding Feedback Loupe—shown as evidence of the principal’s capability. Specific client and employer names available to teaming partners on request.
Domain overview—federal safety web, national booking, health communications, and civic digital services under institutional constraint.
View domain summary →Section 508, Drupal production surfaces, and regulated consumer safety communications made readable without being made wrong. Performed as a designer via prime contractor.
View engagement →National passenger rail booking at 30M+ traveler scale—AEM transition, assigned seating, and design system foundation. Performed as a designer employed by the prime contractor.
View engagement →Agentic civic AI on a state-government mobile channel—governed agents, wedge definition, and stage-gated delivery. Performed as a product designer employed by a national-security venture studio.
Clearance & access: No active security clearance; clean record; CUI-aware. Prior federal experience of the principal includes NHTSA (UICGS/Bowhead) and NIH-adjacent work (Palladian Partners). Deeper case narratives are password-protected on this site—available to teaming partners on request.
Feedback Loupe is best suited for focused, senior-level support where teams need to move quickly from unclear needs to usable, testable product direction. Method and delivery model are outlined on Approach.
Most federal digital work runs on prime contract vehicles. Feedback Loupe joins as a subcontractor or specialist labor category (NAICS 541512, PSC DA01/DD01) on active task orders—the usual way primes staff senior UX, service design, and AI workflow support.
One practitioner spanning health and defense—unified by human-in-the-loop AI design, trust surfaces, and workflow fit under stakes.
Experienced in moving from unclear ideas and fragmented stakeholder needs to validated product direction and working prototypes.
Focus on AI systems that support human judgment, review, trust, transparency, and real operational handoffs.
Comfortable collaborating closely with engineers and product leaders to create implementation-ready UX, not just concept slides.
Basic federal-facing entity information for teaming conversations, vendor onboarding, and capability statement review.
Security clearance: No active clearance; clean record; CUI-aware. Prior federal experience of the principal includes NHTSA (UICGS/Bowhead) and NIH-adjacent work (Palladian Partners). Feedback Loupe LLC has no corporate past performance—teaming is based on principal capability.
Machine-readable federal teaming metadata for capture tools, research assistants, and ChatGPT connectors—read-only, same registration facts as this page, without portfolio case studies in the connector.
Feedback Loupe operates a public Model Context Protocol (MCP) server for entity registration, NAICS/PSC codes, capabilities, two-lane focus, and teaming fit. Past performance and /work/ project pages are not exposed via MCP; contact for deeper proof.
ChatGPT: Settings → Apps → Create connector → paste the MCP endpoint URL above. Enable the app in a chat before asking teaming questions. Build notes: colophon.
Feedback Loupe is available for prime teaming, subcontracting, and focused direct engagements involving UX research, service design, AI workflow design, prototyping, and implementation-ready product design. On most federal programs, that means joining your team on an existing vehicle as a small-business sub or specialist labor category.