In Development

Boat HQ — a yacht that runs itself.

A full-stack operations platform for live-aboard yacht owners. Maintenance schedules, voyage logs, system-by-system manuals, crew, inventory, expenditure, and an AI assistant grounded in the boat's own documentation. One place for everything that keeps the boat moving.

SectorMarine · SaaS
StackNext.js 15 / React 19 / SQLite
AIAnthropic Claude
StatusIn active development
01/The problem

Yacht ownership is a paper chase.

Live-aboard yacht owners juggle binders of manuals, scattered receipts, half-remembered service intervals, a spreadsheet of crew details, and a constant fear of forgetting the one task that ruins a passage. Everything that matters lives in the wrong place.

The off-the-shelf tools are either too generic (notes apps, calendars) or too narrow (anchor watches, chartplotters). Nothing covers the full operational picture — the maintenance, the parts inventory, the voyage history, the crew, the manuals, the spend — in one coherent place. So owners build their own ad-hoc systems and lose hours every season stitching them together.

02/The solution

One platform for everything that floats.

Boat HQ is a complete operations cockpit for a sailing yacht. It models the boat as a real system — 15 categorised subsystems, each with its own maintenance schedule, parts list, manuals, and history. It tracks voyages, crew, inventory, expenditure, seasonal prep, and chandlery. And it has an AI assistant trained on the boat's own documents, so the answer to "when did we last service the saildrive?" is one sentence away — with the receipt.

15 marine systems Voyage log + weather Manuals library Crew + permissions Inventory + chandlery Seasonal prep AI assistant Drive sync
03/What's inside

A real model of the boat — module by module.

Not a generic CRM rebranded for boats. Boat HQ understands what a yacht actually has on board, and what it needs to keep running.

Systems

15 categorised subsystems — engine, steering, rigging, hull, electrical, navigation, plumbing, accommodation, tender, refrigeration, galley. Each carries its own maintenance schedule, parts, and manuals.

Maintenance

Calendar-driven and hours-driven schedules per system. Priority weighting, due-date alerts, completion history, photo evidence, drag-and-drop reordering of the immediate queue.

Voyage log

Route, date, weather, sea-state, engine hours, crew on board, notes. Auto-aggregates total engine hours; surfaces patterns over the season.

Manuals library

Bulk-import PDFs from Google Drive, tag by system, full-text search. Every manual links to the systems that use it, with the AI able to cite specific pages.

Crew

Crew profiles, qualifications, watch rotations, contact details. Per-voyage manifest with sign-on/sign-off, ready for marina paperwork.

Inventory + chandlery

Track spares, consumables, ropes, tools. Tag by system, location on board, supplier. Re-order alerts when stock dips below threshold.

Expenditure

Log spend against system or voyage. CSV import for bank exports. Annual rollups, supplier rankings, "what did we spend on the engine this year" at a glance.

Seasonal prep

Lay-up and recommissioning checklists, repeatable year over year. Drag-and-drop ordering, photo evidence, sign-off when each item is complete.

Knowledge + AI

Boat-specific knowledge base — your manuals, your service history, your notes — searchable via natural language with citations back to the source documents.

The AI layer

Ask Bilbilis. Get answers grounded in your boat's own knowledge.

Boat HQ ships with a custom AI assistant powered by Anthropic's Claude. It's trained on every manual, every voyage log, every maintenance record, and every receipt you've ever logged for the boat. Natural-language in, citations out — never a hallucinated answer.

Service intervals — "When did we last replace the impeller, and what part number is in the bilge locker?"
Manual lookups — "What does the Volvo Penta manual say about prolonged storage in salt water?"
Voyage memory — "What was the engine behaviour last time we beat into a 25-knot SW headwind?"
Provisioning — "How many days of galley supplies for four crew on a passage to Brittany?"
Compliance — "What safety equipment expires before September, and what's the cost to renew?"
Troubleshooting — "Why is the bilge pump cycling every 12 minutes? Walk me through it."
04/How it's built

Modern stack, marine logic.

Boat HQ is a fully-typed Next.js 15 application running on the React 19 server component model, with a SQLite data layer for fast embedded queries and an Anthropic Claude backbone for the AI features.

The frontend uses Radix UI primitives for accessible components, Tailwind for layout, and dnd-kit for the drag-and-drop boards across tasks and seasonal-prep. The data model is hand-built around marine systems — not adapted from a generic SaaS schema — so the application behaves like it understands a boat, because it does.

Frontend · Next.js 15 + React 19 UI · Radix + Tailwind + dnd-kit Data · SQLite via better-sqlite3 AI · Anthropic Claude SDK Types · TypeScript + Zod schemas Auth · bcryptjs + sessions Sync · Google Drive import Hosting · Railway, Docker-packaged
05/By the numbers

Built for a real boat, with real depth.

15
Marine systems modelled
17
Functional modules
42+
Manuals indexed in dev
100%
Typed end-to-end
06/Where it's going

From a single boat to any boat.

Boat HQ is being built and tested against a real sailing yacht — SY Bilbilis, currently moored in Falmouth. The depth of the data model and the AI integration came from solving the actual operational problem first, not from a generic SaaS template.

The next phase is opening it up: multi-boat support, marina integrations, parts-supplier APIs, and a marketplace where qualified surveyors and engineers can plug into the same operational record. The goal is the same as every Layered build — an honest tool that earns its place by being genuinely useful, every day.

Got a tool inside your business nobody else has built?

We'd love to hear about it. Whether it's a marine platform, an internal app, or something nobody's tried yet — we'll tell you honestly if we can help.

Start a Conversation