Apps / semaiASSETS

Coming Soon 1Q 2027

semaiASSETS

Material Master Data Management and Cataloguing

One material catalogue across entities: every item that is bought, stored and used gets one record, written the same way everywhere, so the same item is recognised as the same item in every warehouse. AI is used as a tool to clean up the material data; a person decides every proposal.

For the complete view, open this on a laptop, PC or tablet.

Five catalogue lines that all read BEARING WHEEL, rewritten as standard descriptions from a public reference catalogue: four tapered roller cones and one cup (from our pitch, interface in Bahasa Indonesia)
5catalogue lines reading BEARING WHEEL turned out to be four cones and one cup
17companies registered the same part number for the same bearing in a public reference catalogue
0 → 10of 13 part numbers with a standard description after cleansing the 14-line sample
4weeks for a sample diagnostic that reads your data and changes no system

What semaiASSETS Does

Material master is one record for every item that is bought, stored and used: name, unit, manufacturer's number, material group. Procurement, warehouse and maintenance read the same record. The catalogue is the rule for writing it: a noun that names the item, a modifier that narrows it, then attributes in a fixed order and the manufacturer's number.

Cleansing Workbench

Reads every old line, takes the manufacturer's number out of it, matches it to a reference and writes the standard description from the reference values.

AI as a Tool

AI proposes a noun, a class or a likely duplicate. A cataloguer accepts or rejects each proposal, and every value keeps its source: manufacturer, web page, physical check or model.

Duplicate Clusters

Look-alike records are grouped and compared side by side. The reference values decide, not the look of the text: lines without a reference go to review, never a guess.

Shared Dictionary and Templates

Approved noun and modifier pairs, each with its attribute template in a fixed order, plus units and abbreviations, aligned with international cataloguing standards.

One Catalogue Across Entities

One identity per item, with each entity's own code mapped to it. Every entity keeps its own system, so none has to be replaced to start.

Keeping It Clean

Every new material request passes the template, a duplicate check and an approval flow before a code is created, and data quality is measured the same way every month.

Why semaiASSETS

When each entity writes records its own way, the same item gets different names and different items share one name. The result is double purchases, idle stock, and items that already sit in another entity's warehouse but cannot be seen. Built for groups of entities with many warehouses: Oil & Gas, Mining, BUMD / BUMN.

Same Text, Different Items

Five lines in a public fleet parts catalogue read BEARING WHEEL. Matched to a reference, they are four cones and one cup. Merging them because the text matches would join different items.

Idle Stock Has a Cost

Items in storage keep costing money: tied-up capital, warehousing, insurance and depreciation, at 12% to 30% of their value a year in the published literature.

Transfers Need One Identity

Moving surplus between entities works only when the item in the sender's warehouse and the item the receiver needs read as the same item, even with different codes in two systems.

Many Entities, Many Systems

Entities run different ERPs that will not be merged. A central dictionary with each entity's catalogue and a code map shares the identity of an item, not the system.

How semaiASSETS Works

One real catalogue line taken through every step: collect, take the number, match the reference, write the standard, check duplicates, merge or retire, keep it clean. A line without a reference goes to the review list and is completed from the manufacturer. AI proposes; the cataloguer decides.

How semaiASSETS WorksHow semaiASSETS Works (1/3)How semaiASSETS Works (2/3)How semaiASSETS Works (3/3)

For the complete view, open this on a laptop, PC or tablet.

Open the animated diagram

Solution Architecture Overview

The inputs on the left, the modules semaiASSETS adds in the centre (clean, keep it clean, catalogue), each entity's own system and the outputs on the right, the people along the bottom, and the questions still open, each numbered.

semaiASSETS solution architecture overview: material master extract, surplus list and coding standard as inputs; reference catalogues and manufacturer data; inside semaiASSETS the cleansing workbench, AI assist, duplicate clusters and review list, the material request, duplicate check, approval flow and data quality score, and the shared dictionary, shared catalogue, cross-entity stock and decision log; each entity's ERP with mapped codes; outputs a clean catalogue load and a diagnostic report; three open questions

For the complete view, open this on a laptop, PC or tablet.

One Catalogue Across Entities

Each entity keeps its own system; its codes are mapped to one identity per item. Surplus in one warehouse is then seen by another entity before anyone buys, and can be transferred or loaned under each entity's rules.

One Catalogue Across EntitiesOne Catalogue Across Entities (1/3)One Catalogue Across Entities (2/3)One Catalogue Across Entities (3/3)

For the complete view, open this on a laptop, PC or tablet.

Open the animated diagram

Mockups

Concept screens. The items and their values are real, from a public-domain fleet parts catalogue matched to a public reference catalogue; entities, stock and request numbers are examples. Interface in Bahasa Indonesia.

Cleansing workbench: 14 old catalogue lines with their part number, reference record, standard description, status and action; ten matched, three missing attributes, two to review, one pair to merge
Cleansing WorkbenchEach old line is read, its part number taken, matched to a reference and rewritten. Lines without a reference go to review.
Duplicate cluster: five records with the same description BEARING WHEEL compared side by side; the reference values show four cones and one cup, decision: split into five identities
Duplicate ClustersCandidates compared side by side; the decision comes from the reference values, not from the look of the text.
Data quality score: before and after charts for standard descriptions, identities behind BEARING WHEEL and records for one part number, with a table of measures and their status
Data Quality ScoreMeasures before and after, counted from the 14-line sample; in a real project, for every entity's whole catalogue.
New material request: noun, modifier and attributes chosen from the dictionary, the description written by the system, and a duplicate check before a code is created, finding the item already exists
New Material RequestNoun and modifier only from the dictionary; the system writes the description and checks for duplicates before any code is made.
Dictionary and template: approved noun and modifier pairs, each with its ordered attribute list and which attributes are required, and the description the template writes
Dictionary and TemplatesApproved noun and modifier pairs, each with its attributes in a fixed order.
Cross-entity catalogue: one item identity with each entity's own code and stock; one entity needs 4, another holds 12 in surplus, and a transfer is proposed
Catalogue Across EntitiesOne item identity, with the code and stock in every entity, and a transfer proposal under the applicable rules.
Approval flow: six steps from request to code creation, a queue of requests with their age, and the automatic checks for one request
Approval FlowEvery new item passes the same steps before it gets a code; the automatic checks reject what does not fit the dictionary.

Let's Build #SmarterEnterprise

Start with: NDA & Joint Planning Workshop