For yacht builders

Owner support, on tap.
Factory-authorized. In 30 days.

A branded AI engineer that knows every model you build, quotes your service intervals, and answers your owners' questions with the depth of a chief engineer, all day, every day. Deployed on your domain. Trained on your data.

Your owners have questions. Your dealers are booked out for weeks.

Your marketing team put together beautiful brochures. Your dealer network is doing their best. But when a captain's engine throws a fault code at 3 AM in the Bahamas, they're on Google, on forums, and on YouTube. Not on your site.

1

Owners search alone

60% of owner-support questions never reach a dealer. Captains fix things off Google, guess, or wait. Your brand is invisible in the moment that matters.

2

Dealers are stretched

Your service network handles the hard cases. It should not be answering "what's my 250-hour service" or "which Racor fits this engine." Free them up.

3

Bad answers hurt your resale

Every wrong forum post ("just deleted my EGR, runs great") becomes tomorrow's failure. And tomorrow's insurance claim. And tomorrow's negative review.

What we build for you.

A factory-authorized owner-support surface deployed on your brand, connected to your service data, live in 30 days.

Focus owner support dashboard for a Focus 3X yacht, brand-skinned, showing service intervals, warranty status, and factory-authorized chat.
Live surface. Every element (skin, sidebar, tiles, chat) is what your owners would see. Open it →
Answer quality · real example
Focus 3X owner support
Open the live demo →

"I have a Focus 3X, 600 hours in, EGT is climbing at cruise. What am I looking at?"

Raw-water pump impeller. Common at 500 to 600 hours on the Yanmar 8LV, neoprene vanes fatigue in salt water and cooling flow drops before you see any visible failure. Replace with Yanmar p/n 129470-42532. While the pump is open, check the cam plate for scoring. Verify EGT drops back to baseline at cruise after replacement. If it doesn't, next check is the aftercooler tube stack for scaling.
Answered by the actual /focus surface. Not a mockup. Try it yourself.
~1.4s
Time to first token
30 days
Ingest to launch
7
Languages supported
24/7
Owner coverage

How the engagement runs.

1

Week 1: ingest

Send us your model line, brochures, and service manuals. We build the knowledge base and per-model service intervals, failure patterns, and part cross-references. You review, we cross-verify.

2

Week 2-3: brand

Your logo, your colors, your voice guidelines. Deployed at ownersupport.yourbrand.com or embedded in your site. Every conversation feels like a factory-authorized channel.

3

Week 4: launch

Live for your owners. Weekly analytics: top questions, top failure patterns, geographic hotspots. Data your product team has never had before.

Beyond chat.

A chat window is the surface. Underneath, the engineer plugs into your dealer network, your CRM, and the languages your owners actually speak.

Escalation to your dealer

When a thread crosses a confidence or safety threshold (parts request, warranty claim, damage assessment), the engineer opens a case in your dealer network CRM via signed webhook. Full transcript, vessel context, GPS if consented. Your dealer picks it up warm, not cold.

Multilingual out of the box

Owners chat in English, Italian, French, Spanish, German, Portuguese, or Dutch. The engineer replies in the same language, quotes your service intervals in local units, and references the same source-of-truth catalog. No separate model per language, no separate ingest.

Analytics your product team wants

Weekly digest of top questions, top failure patterns, and geographic hotspots. Feeds directly into your product roadmap and dealer training. Data your warranty team has never had at this granularity before.

How it's packaged.

One-time setup covers the buildout. Annual license covers hosting, tuning, monthly updates, and priority support. No per-seat, no metered token bills. Pricing scales with model-line size and depth of ingest. We'll size it on the call.

Setup

Custom
Scoped per builder
  • Full model-line ingest
  • Service intervals + failure patterns per model
  • Brand skin, custom voice tuning
  • Deployed at your subdomain
  • Blind eval set for your brand

License

Custom
Annual, all-in
  • Unlimited owner conversations
  • Monthly knowledge base updates
  • Weekly analytics on top questions
  • Model-level fault-code decoding
  • Priority support + dedicated Slack

Security and your data.

Your OEM data stays yours. Straight answers on the questions procurement always asks.

Tenant isolation. Every OEM knowledge base is scoped by tenant ID. Your service manuals and brochures never mix with another builder's corpus, and cross-tenant queries are impossible by design (RLS + scoped API keys). If you end the engagement, we return your data and delete our copy within 30 days.

Training on your data. We do not train foundation models on your content. Your corpus powers retrieval only. If we ever change that, it requires a signed addendum from you first.

Infrastructure. Production-hardened API. Rate limiting per-IP and per-session, timing-safe admin auth, CORS locked in production, signed webhooks. Data hosted in US-East by default, EU (Frankfurt) available on request for GDPR-scoped builders.

Compliance today. DPA available on request. Standard MSA with confidentiality + IP assignment clauses. SOC 2 Type I in progress (target Q2). If your procurement needs SOC 2 Type II before signing, we can discuss timing.

Content rights. Every document in your knowledge base is either OEM-approved, publicly licensed, or fair-use with a cited source. No copyrighted manuals scraped from forums.

Liability & answer accuracy. Every answer carries a source citation your owners can click. The MSA includes standard hold-harmless language for owner interpretation of advisory answers. E&O coverage details in the contract.

Questions procurement always asks.

Who owns the model and the knowledge base?

You own your content and the derived knowledge base. We own the platform code. If you leave, you get an export of your corpus and any structured intervals + failure patterns we derived from it. The AI runtime itself stays with us.

Can we self-host later if we want to?

Yes, under a separate license tier. Self-hosted deployments run on your infra with our runtime image, and you continue to get monthly knowledge updates. Most builders start hosted with us for the first 12 months to move fast, then decide.

What data do we need to hand over?

Service manuals per model line, sales brochures, service bulletins if you have them, standard interval schedules, and a technical contact who can answer factory questions during ingest. That's the minimum. Fault-code decoders, warranty databases, and dealer network directories improve depth but aren't required to launch.

What happens when the engineer gets an answer wrong?

Every reply carries a source citation and an inline "not helpful" signal owners can tap. Flagged threads land in a weekly review queue that we triage together with your team. Systematic misses feed back into the knowledge base within the next monthly update cycle. Safety-critical topics get a lower confidence threshold and route straight to your dealer network.

How is this different from putting ChatGPT on our site?

Generic LLMs guess with confidence. This engineer is scoped to your corpus, refuses to guess outside it, cites its sources, and knows your model line down to serial ranges and part numbers. It also handles the marine-specific state machine (vessel identification, fault-code decoding, escalation to dealer) that a generic chat wrapper doesn't.

Ready to see it work on your brand?

Bring us your five flagship models. We'll show you a working POC in 3 to 5 days at no cost. If it moves you, we build the full line in 30 days.

Book a 30-min call →