Longsplice
For maintenance & engineering organisations

One line through everything you know.

Longsplice is the intelligence layer above the systems you already run. It splices your CMMS, your ERP, SharePoint, Google Drive, OneDrive, Dropbox, change registers and project tools into one operational memory, then answers questions, finds repeat failures and warns you early.

CMMS · Maximo ERP · SAP SharePoint · OneDrive Google Drive · Dropbox monday.com Email · PDF · CSV Longsplice
Splices with Maximo SAP SharePoint OneDrive Google Drive Dropbox monday.com Email CSV · PDF + anything with an API
The problem

Knowledge leaves. Data hides. Failures repeat.

Your organisation already owns the answer to almost every operational question it faces. It just can't find it.

The fix is in someone's head

When a senior engineer retires, thirty years of "we tried that in 2011, here's why it failed" walks out the gate. The work order history stays behind, unread, in free-text fields nobody searches.

Six systems, no memory

The CMMS knows the failure. SharePoint holds the investigation. The photos sit in a Google Drive folder someone shared once. The change register has the fix, and the project tool has the schedule it slipped. No system knows the whole story, so no person does either.

Patterns surface too late

A repeat defect looks like routine noise until the fourth event damages equipment. Executives find out from last month's report, after the docking window has closed.

The product

Five modules on one knowledge tree

Every record, document, drawing and email is linked into a single tree of assets, work, changes, decisions and people. The modules are different windows onto the same memory.

Maintenance Intelligence

  • Repeat defect detection
  • Asset health scoring
  • MTBF trends and reliability growth
  • Spares and stockout risk

Engineering Intelligence

  • Change impact analysis
  • Drawing relationships
  • Design history and lessons learned
  • Compliance tracking

Organisational Knowledge

  • Plain-language search
  • Answers with cited evidence
  • Expert finder
  • Knowledge capture for retiring staff

Project Intelligence

  • Schedule risk and slip velocity
  • Action and decision logs
  • Contractor performance

Executive Intelligence

  • Daily operational briefing
  • Emerging risk feed
  • Cost and reliability trends
  • Fleet and site health

Evidence or it didn't happen

Every answer cites the work order, document or change it came from, with a link back to the source system. No citation, no claim.

The knowledge tree

Every record finds its branch

Underneath the modules is one tree: fleet to vessel to system to component. Everything the connectors ingest is filed onto the branch it belongs to, whichever system or drive it came from.

One branch of the tree

CV-02's raw water pump and everything the tree holds about it. Dashed links were made by Longsplice; no source system connected these records.

linked in the source system smart-matched by Longsplice Coral-class fleet CV-02 seawater cooling DG1 raw water pump WO-25-0334 · Maximo intake photos · Google Drive MMS-PROC-114 · SharePoint ECR-2025-018 · change register docking task · monday.com decision D-14 · meeting minutes

Smart matching

Most records arrive with no links at all. A work order in Maximo, a photo folder in Google Drive, an email thread and a monday.com task can all describe the same pump without ever referencing each other. Longsplice matches them on what the records already carry: asset codes, drawing numbers, serial numbers, dates, names and the language of the text itself. Matched items join the same branch with a confidence score and the evidence for the match.

Decisions are records too

The most expensive knowledge an organisation loses is why. Decisions get made in meetings and inboxes, then scatter. The tree holds each one as a node of its own: what was decided, by whom, on what evidence, and what it affects. When the question comes back in three years, the decision and its reasoning are one hop from the asset.

How it works

An overlay, not a migration

Longsplice sits over the top: your systems stay the systems of record, and everything they hold starts working together.

Connect

If it has an API, it splices in. One universal connector speaks to any REST or GraphQL system with a config file and no code: your CMMS, your ERP, monday.com, SharePoint, Google Drive, OneDrive, Dropbox. CSV, PDF and email ingestion cover everything else. First system live in days.

Splice

Records and documents are normalised, embedded and filed onto the knowledge tree. A work order that mentions a drawing becomes a real relationship you can traverse, and records that never mention each other are smart-matched onto the same branch.

Answer

Ask anything in plain language. Nightly jobs work the other direction, surfacing repeat defects, deteriorating assets and schedule risk before anyone asks.

Your data

Too big for anyone to read. That's the point.

There is nothing to create and nothing to migrate. Longsplice works on the history your organisation already owns, and that history passed human scale years ago.

Work orders
182,000
11 years of free-text history
Documents
64,000
reports, drawings, minutes, manuals
Email
9 years
threads and attachments
Systems & drives
14
none of them talking

Volumes like these are ordinary for a mid-sized operation. At one record a minute, the work orders alone are more than a year and a half of reading. Nobody has done it and nobody ever will. Longsplice reads every record on the way in, matches it into the tree, and re-reads the whole graph nightly, so the full history sits behind every answer, every match and every warning.

Pricing

Three packages, pricing TBD

The commercial model is being finalised. Annual contracts in AUD; asking questions will never be metered.

Foundation

TBD / year
  • Single site, 25 users included
  • Universal connector, unlimited systems
  • Maintenance + Knowledge modules
  • Managed SaaS (AU region)

Operations

TBD / year
  • Multi-site, 100 users included
  • + Engineering and Project modules
  • Managed SaaS or dedicated instance
  • Priority support

Enterprise

TBD / year
  • Fleet scale, 250 users included
  • All modules, named TAM
  • On-premises and air-gapped options
  • Bring your own LLM endpoint
Security

Built for buyers whose security teams read everything

Sovereign by design. AU-region SaaS, dedicated instance, or fully air-gapped on-premises with a local model. Defence customers never send a byte offshore.
Your permissions, honoured. Document ACLs sync from source systems. If someone can't open it in SharePoint, Longsplice won't retrieve it for them.
No training on your data. Contractually and technically. The model is plumbing; your knowledge graph is yours and stays that way.
Isolation enforced in the database. Row-level security per organisation in PostgreSQL itself, not just the application. Every AI call and admin action is audit-logged.
DefenceMaritimeMiningUtilities RailOil & GasManufacturingInfrastructure

Your organisation already knows.
Longsplice remembers.

Overview

Open intelligence across connected systems, worst first. Nightly run completed 03:10.

Critical insights
1
unchanged this week
Fleet health
78.2
▾ 0.6 vs last month
Open P1 work orders
1
WO-25-0501, day 9
PM compliance
91%
▴ 2 pts this quarter

Class-wide DG cooling water blockage pattern

critical

Four corrective work orders across CV-02 and CV-05 share the COOL-BLOCK signature: marine growth fouling seawater intakes, escalating from alarms to protective shutdowns and a damaged impeller ($14.6k). Consistent with a class-wide intake design susceptibility in tropical waters. ECR-2025-018 raised; four vessels unassessed.

WO-24-0187WO-24-0912WO-25-0334WO-25-0281ECR-2025-018gdrive: intake photos

CV-05 health deteriorating three weeks running

warning

Corrective-to-preventive ratio rising and two PMs overdue. Trend driver is the seawater system group. Recommend bringing the intake inspection forward.

CV-05health model v1

Docking-window milestone slipping on the Fleet Docking Program

warning

The CV-03 intake assessment milestone is forecast 22 days past baseline on the monday.com Fleet Docking Program board. It gates ECR-2025-018 work in CV-03's September docking window; at current slip the window is missed.

monday: CV-03 intake assessmentECR-2025-018

Steering HPU spare seal kits below reorder point

warning

Two kits on hand, 11-week lead time, and WO-25-0501 is likely to consume one. A second HPU event before November would mean a stockout on a criticality-5 system.

WO-25-0501stock: SK-HPU-09

Corrective work orders per quarter

CV-02 diverged from the fleet average four quarters before the impeller failure.

CV-02Fleet average

Cost per COOL-BLOCK event

The pattern got expensive at event three. Early detection pays for the year.

Ask Longsplice

Plain-language questions over every connected system. Answers cite their sources; no citation, no claim.

Assets

Coral-class fleet, scored nightly from failure trend, corrective ratio, overdue PMs and defect recurrence.

CodeVesselHealth12-month trendState

Engineering changes

Every change with its blast radius: affected assets, drawings and procedures traced through the knowledge graph.

ECR-2025-018 · Seawater intake grate redesign and antifoul upgrade

in assessment · day 38

Raised by L. Tran off the class-wide DG cooling blockage pattern. Risk rating medium. Docking windows for CV-03 and CV-06 close in September; assessment must complete by 15 August to make them.

Affects assets
All six Coral-class vessels
Affects drawings
CC-DWG-3305 rev B
Affects procedures
MMS-PROC-114 rev C
Source evidence
MMS-REP-2025-041 + 4 WOs

Daily briefing

Composed each morning from the last 24 hours of insights, changes and risks. Delivered in-app and by email.

Meridian Marine Services · Sunday 19 July 2026

Needs a decision. ECR-2025-018 (intake grate redesign) has been in assessment 38 days. Four vessels remain exposed to the cooling blockage pattern that damaged CV-02's DG1 impeller ($14.6k). Docking windows for CV-03 and CV-06 close in September; the assessment needs an owner and a date.

Watch. CV-02 steering HPU pressure fluctuation (WO-25-0501) is still open at priority 1, nine days in progress, and spare seal kits are below reorder point with an 11-week lead. CV-05 health trend remains deteriorating for a third week.

Good news. CV-04 preventive completion is at 96% for the quarter and its health score is up 4.1 points since April. Fleet PM compliance is the best it has been this year.