System of intelligence
WoyCore AI
The AI core: the knowledge graph, a verified knowledge base with source citations, and tool-constrained AI agents that answer, draft and validate.
WoyCoreAI platform for fleet maintenance and compliance
WoyCore connects equipment, spare parts, documents, regulations and maintenance history into one knowledge graph. Its AI agents answer with the exact source, draft the paperwork for your confirmation and check that every closed job is backed by evidence, in the office and, with the onboard node, on board even when the satellite link is down.
WoyCore AI AssistantPublic beta Q4 2026
Capabilities are introduced in stages – see the roadmap.
01The problem
Most planned maintenance systems were designed as digital logbooks for record keeping and reporting. They depend on manual entry, they often accept “Done, checked OK” as proof, and they record equipment as a list rather than as a network of dependencies. Meanwhile the knowledge a crew needs is spread across tens of thousands of pages of manuals, drawings, circulars and procedures.
The answer exists, somewhere in a 1,000-page parts catalogue, a service letter from last year or a company circular. Finding it takes hours the crew does not have.
A job marked “Done” without measurements, test results or the parts used tells the office nothing, and it will not convince an inspector.
A language model that does not know your ship’s layout can invent a part number or blend international, local and company rules into one confident, wrong answer.
Cloud tools stop working where the satellite link drops, which is exactly where the crew needs them.
Handovers are written in a hurry. Chronic issues and hidden defects go home with the person who knew about them.
02The platform
WoyCore is built from the ground up for machine reasoning and verifiable answers. At its centre is a knowledge graph of your fleet: every vessel, system, component and spare part, every document and revision, every regulation and where it applies, every job, defect and certificate, linked by stable identifiers and traceable to its source.
System of intelligence
The AI core: the knowledge graph, a verified knowledge base with source citations, and tool-constrained AI agents that answer, draft and validate.
System of record
A modern, lightweight planned maintenance system designed around evidence: equipment tree, planned maintenance, defects, requisitions, certificates, SMS document library, findings and corrective actions, digital checklists.
03How it works
Manuals, SMS procedures, circulars, certificates, forms and PMS exports flow in through self-service upload. Every document is checked, classified and approved before it can be used as a source.
The knowledge graph links each item to the vessel, system and equipment it belongs to, to the regulations that apply and to the revision that is current. One physical unit, one identity, across every module.
Agents ask the graph what applies first, then search only inside the verified, applicable sources. You get an answer with the exact clause, or a ready draft of a work order or requisition.
Nothing is approved or closed by AI. A person confirms every change, and a job is not ready for closure until the measurements, test results and evidence required for its type are recorded.
04Capabilities
Equipment hierarchies, dependencies, spare parts, documents and regulations in one connected model. WoyCore knows that a pump cools the main engine, so it warns about the consequences, not just the repair.
Every substantive answer cites the document, revision and clause it comes from. When the data is not there, WoyCore says so and tells you what it needs.
Work orders, requisitions and forms are drafted with exact part numbers and linked procedures. Unknown measurements stay blank. A person confirms every change.
Mandatory facts per job type, sequence and time checks, comparison with limits and history, in-app photo evidence with re-use detection. Formal closure gets flagged.
A dedicated onboard node keeps the full functional scope, local AI and the vessel’s knowledge working for as long as the link is down, and synchronises when it returns.
Built on established maritime standards for documentation, equipment coding and supply cataloguing, instead of proprietary formats.
05In practice
01
The situation
The engineer reports low pressure on the fresh water cooling pump and asks what to do.
What WoyCore does
The graph shows the pump cools the charge air cooler of the main engine. The repair procedure is pulled from the pump’s own manual, with the clause and page reference.
The result
The engineer gets the repair steps and the warning that matters: the main engine’s charge air cooling is affected, with the engine manual’s guidance on load cited.
02
The situation
A seal for a separator is out of stock. Can the one from another pump be used?
What WoyCore does
Interchangeability is checked against physical properties recorded in the graph: dimensions, material, pressure rating. Stock on board is checked through the PMS.
The result
A compatible part number, its specification and the quantity on board, instead of a guess based on similar wording.
03
The situation
The engineer opens the original engine manual to prepare a piston overhaul.
What WoyCore does
Before answering, WoyCore checks whether a later service letter modifies that procedure, and finds one.
The result
The answer shows both the original clause and the change, with the service letter reference, so the outdated procedure is not followed.
06For ship managers and operators
Less administrative load
for crew and shore staff.
Lower exposure at inspections
port state control, vetting, class and audits.
Fewer issues lost in closed jobs
deviations become managed defects.
07Deployment
A dedicated, isolated instance for your company, in the cloud or on your own infrastructure. Fleet-wide knowledge, company documents, integrations and heavy processing.
An autonomous instance on the vessel with the knowledge relevant to that ship and a local AI model. Works with no connectivity and synchronises in short, resumable sessions.
A public, cloud-based AI Assistant for marine engineers, officers, cadets and instructors, answering from openly published regulatory and reference sources. Public beta in Q4 2026.
08Roadmap
Research and the architecture baseline are complete, and development of the first release is under way. Capabilities arrive in stages, each one tested on real questions before it ships.
Now Q4 2026
Cited answers from openly published maritime regulatory and reference sources, for marine professionals.
Next
A company instance with its own knowledge base, WoyCore PMS and the first AI agents, piloted on real vessels.
Then
Full capability on the vessel without connectivity, with a local AI model and signed knowledge updates.
Later
Mail and event-driven checks, department agents, condition monitoring through a one-way gateway.
09Contact
Questions about WoyCore, the public beta of the WoyCore AI Assistant or the platform for your fleet? Write to us.