ShipLock
BuildingA client delivery protection system for dev studios — scope locking, AI drift detection, signed client review links, and a full audit trail so “that's not what we asked for” becomes impossible.
The brief. Seven months into a client engagement, our team hit a wall: a 40-page requirements document, four weeks to deliver everything, and no way to prove what was agreed versus what was added. The client said “that's not what we asked for.” We had no system to say otherwise.
The real problem. Nobody locked anything. Requirements lived in WhatsApp threads and meeting notes. Demos happened but approvals never did. Scope expanded by roughly 300% — invisibly, incrementally, with no paper trail. The failures were structural, not technical. There was no system enforcing alignment between what was agreed and what was expected.
The approach. I built ShipLock from that pain — not imagined problems. It sits on top of an existing workflow and enforces one principle: nothing exists unless it is documented, acknowledged, and traceable. AI extracts and locks requirements. Demos route through a signed approval pipeline. Scope changes get flagged and costed before a single line of code is written. The audit trail is the product.
The app, running live
The screens below are the real ShipLock app, running live on a seeded demo you can click through — open a requirement, switch tabs, move around the sidebar. The projects, clients, and numbers are fictional; everything else is the production build.
The delivery command centre: what is shipped, what is blocked, and by whom. "Blocked by client" reads as fact, not accusation — so when an exec asks why a project slipped, the answer is already on the screen. This is the real production dashboard running live; only the projects and numbers are fictional.
Upload the brief and the AI extracts every requirement into a trackable item with a unique ID. The client approves or disputes each one, and the 48-hour rule turns silence into consent — no response is auto-approval. Live requirements screen, fictional data.
Paste any meeting note or email and the AI scans for drift, flags anything outside approved scope, and forces a timeline-impact estimate before work begins. The client has to acknowledge the impact — no acknowledgement, no work. Real screen, fictional data.
Every client-facing send starts a timer: green, yellow at 24h, red at 48h, auto-escalation at 72h. Responsiveness becomes a visible, shared record instead of a chase over email. The real tasks board, running live with fictional data.