ExitProof

Inspiration

Employee offboarding is usually scattered across an HR system, IT tickets, spreadsheets, inboxes, and somebody's memory. That works until a former employee still owns a cloud billing contact, payout notification, organization-owner role, production token, or company data. The costly failure is not merely forgetting a checklist item; it is having no defensible proof that control changed hands.

What it does

ExitProof turns a plain-language departure brief and a reusable control template into an auditable closure plan. Its Xano AI agent adds an executive risk analysis that separates basic account removal from consequential controls such as billing ownership, payouts, tokens, keys, data transfer, and organization-owner permissions. Every control gets an owner, deadline, risk level, and evidence requirement. Final certification is technically blocked until every control is resolved.

How we built it

Xano is the product's control plane, not a storage add-on. Its AI agent analyzes the brief and names the most consequential risks. Its relational backend stores departure cases, submitted access-control templates, and audit events. Its transactional API creates the plan and audit record together. Its resolution workflow attaches evidence and recalculates readiness. Its certification endpoint rejects incomplete exits at the backend.

Codex was used to design and implement the frontend, model the data and API contract, and validate the XanoScript structure. Xano Agent created the live workspace resources; the published OpenAPI contract and end-to-end API tests helped us find and repair incomplete endpoint logic before connecting the frontend.

Challenges we ran into

The hardest challenge was keeping AI useful without letting it become the enforcement layer. The model can identify risks and recommend controls, but only deterministic backend rules can prove whether each control was actually closed. We also had to keep the case, submitted controls, and audit trail consistent across a multi-step workflow.

Accomplishments that we're proud of

• A working AI-assisted offboarding flow backed meaningfully by Xano.
• Evidence-backed controls with owners, deadlines, and risk levels.
• A backend certification gate that refuses incomplete departures.
• An append-only audit trail that makes every decision defensible.
• A public application, API, repository, automated tests, and concise demo.

What we learned

The best AI workflow separates judgment from authority. AI is excellent at surfacing overlooked risk; the control system must remain deterministic, auditable, and difficult to bypass.

What's next

Connectors for identity providers, cloud platforms, source control, finance tools, and HRIS products could discover access automatically. Policy templates could adapt controls by role, jurisdiction, and company risk tolerance while preserving human verification for consequential actions.

Built With

• codex
• next.js
• react
• shadcn
• tailwind-css
• typescript
• xano
• xano-ai-agents
• xanoscript