Jeremy Neely Text Jeremy

Security & Compliance · Prototype

AME CMMC

First stage of an AI-assisted CMMC readiness platform; AI drafting is designed, not built.

Built for: AME staff who assess defense-contractor clients

The framework browser showing a Level 2 requirement and its objectives.
The framework browser showing a Level 2 requirement and its objectives.

The problem

CMMC readiness means reading 110 Level 2 requirements and 320 assessment objectives against a client's security documents. By hand that is days of expert reading per client, and the reasoning trail is easy to lose.

What I built

I built the first stage of an internal platform for assessing defense contractors against CMMC: the database, sign-in, a guide importer and a framework browser. In the full design, AI drafts a status for every objective and an assessor accepts or changes each one.

How it works

  1. Import the official assessment guides as versioned data.
  2. Staff sign in and browse requirements and objectives.
  3. Planned: upload client documents and tag evidence.
  4. Planned: AI drafts findings with cited evidence for review.
  5. Planned: produce readiness, SPRS and proposal reports.
How AME CMMC is wiredOnly the first stage is built. Every AI call is recorded, and a staff member confirms each finding.
Comes inInside AME CMMCGoes out Assessment guides official, versioned Client documents planned Staff sign-in admin-seeded accounts Framework browser 110 reqs, 320 objectives Evidence tagging planned AI-drafted findings planned, staff confirm Readiness report planned, PDF AI activity log tokens, cost, errors

Built with

  • TypeScript
  • Fastify
  • React
  • Vite
  • PostgreSQL 17
  • TanStack Query
Text Jeremy about this Next: Shadow AI Scan and Reporter All 38 tools Texting opens your Messages app with a note already written. He replies himself.

More of Jeremy's work

All 38 tools, by category