Cut shuttle fleet fuel
20–30% without
buying new boats.
A 24-hour assignment solver for offshore wind CTV operators and harbor launch services. Plugs into your existing booking flow via API. No new hardware, no rip-and-replace.
Live deployment
Deployed at a harbor launch operator in Southeast Asia. 30% fuel reduction in a 12-month backtest. Live since 2026; early results tracking similarly.
You're leaving money on the water.
Fuel is the largest controllable cost on a small-vessel shuttle fleet — typically 15 to 25% of opex. Every trip that backtracks to base, every long-haul handed to a thirsty older hull, comes off the bottom line.
Dispatchers can't solve a 24-hour assignment problem in their head. So most operations dispatch greedily — next available boat — and absorb the cost as the price of doing business. It isn't.
An API call between
your booking and the boat.
BunkerCut runs as a hosted service. When a reservation comes in, you call the solver and get back the vessel that minimizes total fuel burn across the next 24 hours — accounting for trips already on the books and the efficiency of every hull in your fleet.
Connect your booking system.
A few API endpoints. Reservations in, fleet metadata once. We integrate alongside Helm CONNECT, custom dispatch tools, or spreadsheets via webhook.
Every booking triggers a solve.
Mixed-integer optimization across a 24-hour horizon. Trip chaining sequences pickups so a boat finishing near job B is dispatched to the next job at B.
Assignment back in milliseconds.
Your dispatcher sees the recommended boat. Override anytime — the solver re-plans around manual decisions on the next call.
{
"booking_id": "BK-3144",
"pickup": { "lat": 25.123, "lng": 56.341,
"time": "2026-05-08T07:40Z" },
"dropoff": { "lat": 25.072, "lng": 56.298 },
"passengers": 8
}{
"vessel_id": "MV-04",
"depart_from": "anchorage_zone_2",
"horizon_fuel_delta_l": -184,
"rationale": "chains with BK-3119 dropoff"
}- →Trip chainingSequence assignments so a boat finishing a drop-off near pickup B is dispatched to the next job at B — instead of sending another boat from base.
- →Fuel-efficiency-aware allocationPrefer efficient hulls for long trips. Reserve thirstier boats for short hops where the difference is negligible.
- →24-hour horizonThe solver plans against the bookings already on the schedule, not just the trip in front of it. Every dispatch is globally aware.
One operator. Twelve months of bookings. Real numbers.
We backtested the solver across 12 months of historical reservations from a harbor launch operator in Southeast Asia (~$2M annual revenue, mid-size fleet). Then deployed it live. Early results are tracking the backtest.
Anonymized one-pager available on request — fleet composition, booking volume, baseline fuel curves, post-deployment month-on-month delta.
Caveat we'd rather say upfront: one customer, early production data. We're not claiming "proven at scale" — we're claiming "the math is straightforward, the backtest is honest, and the live deployment is tracking."
Built for two specific operators.
If you run reservation-driven trips with a mixed-efficiency fleet of 5 to 50 boats, and fuel is your top controllable line — this was built for you. If you don't, it probably isn't.
Offshore wind CTV operators
Fleets of aluminum catamarans servicing offshore wind farms. North Sea, Taiwan Strait, US East Coast, Korea, Vietnam. Buyer is the COO or Head of Marine Operations. Under emissions pressure from utility customers.
- ·Crew transfer between port and turbines
- ·5–50 boats per fleet
- ·Helm CONNECT alongside, not replaced
Harbor launch & bunker tanker operators
Major bunkering ports — Fujairah, Rotterdam, Gibraltar, Houston, Hong Kong. Shuttles between shore and vessels at anchorage. Buyer is the GM or Operations Manager.
- ·Crew, supplies, personnel to anchorage
- ·8–30 boats per fleet
- ·Spreadsheet or whiteboard dispatch today
We don't replace what you already run.
Large fleet platforms target deep-sea commercial shipping. Small-vessel multi-stop assignment is a structural gap. We sit in that gap — and integrate with everything you've already paid for.
Not a fleet management ERP
Helm CONNECT, AMOS, and similar still own scheduling, crew, and compliance. We call into them; we don't replace them.
Not a maintenance system
Engine hours, dry-dock cycles, parts inventory — those stay where they are.
Not a routing or nav tool
We don't pick the route between two points. We pick which boat does which trip.
Not a vessel tracking platform
AIS, GPS, telemetry feed in if you want richer signals. They don't have to.
Priced against the savings we deliver.
- Step 01Paid pilot.Fixed-fee, time-boxed. We integrate with your booking system, backtest against your last 12 months of dispatches, and run live for 60 days. Pilot pricing is scoped to your fleet size and integration. Final number is determined after a short discovery call. We don't engage if the math doesn't work for both sides.
- Step 02Production subscription.Monthly fee tied to fleet size. Optional shared-savings clause: a portion of measured fuel reduction, capped, based on a measurement methodology agreed at pilot kickoff.
Built by people who've actually run boats.
BunkerCut started inside a small-vessel operation, not a tech accelerator. The team combines optimization research (mixed-integer programming, vehicle routing) with hands-on harbor operations — people who've sat in dispatch rooms, looked at whiteboards, and seen exactly where fuel disappears.
We're deliberately narrow. We solve one problem — assignment over a 24-hour horizon — and we solve it well. We're not a platform. We're not a fleet-of-the-future story. We're a solver behind an API, paid back by the fuel you stop burning.
Tell us about your fleet. We'll come back with a number.
No deck pitch, no discovery loop. We'll ask about your fleet composition and current dispatch flow, then estimate the savings the solver would have generated against your last 12 months of bookings. You decide if it's worth a pilot.
- ✓No newsletter, no follow-up sequences.
- ✓NDA on request, before you share data.
- ✓We'll tell you if you're not a fit.