Manufacturing & Industrial
Industrial Parts Matching API
A maintenance engineer types a part number from a label into your search box. If you do not have that exact number, you lose the order — even when you stock three equivalents.
- Industrial distributors
- MRO marketplaces
- Procurement software
- Equipment manufacturers
- Maintenance platforms
What your customers would see
example interfaceMockupIllustrative mockup of the interface we would build. Figures shown are examples, not results from a live deployment.
The problem this solves
Industrial parts carry a number from whoever made them, a different number from whoever distributed them, and often a third from whoever maintains the equipment it goes into. The same bearing is listed four ways across four catalogues. A buyer holding a greasy label with a worn number has no practical way to discover that the item in stock down the road is the same part.
Distributors feel this as lost revenue and as wasted counter time. A search for an unstocked number returns nothing, the buyer phones a competitor, and the order goes elsewhere despite three equivalent items sitting on the shelf. Where cross-references do exist they are usually a spreadsheet one experienced person maintains, undocumented and unshared, which becomes a serious problem the week that person retires.
Building this properly means accepting that equivalence is not binary. Two parts can be dimensionally identical and differ in temperature rating, so they are interchangeable in one application and dangerous in another. A matching service that returns a single “equivalent” flag is misleading. The useful output states the type of equivalence, the specifications that differ, and the confidence — and leaves the fitness decision with the engineer who knows the application.
Who buys this
Industrial distributors, MRO marketplaces, Procurement software, Equipment manufacturers, Maintenance platforms
What the API returns
What comes back
- Candidate equivalents for a part number across manufacturers and distributor catalogues
- Equivalence type: direct replacement, functional equivalent, dimensional match, or partial
- Specification comparison showing exactly which attributes match and which differ
- Confidence per match with the evidence that produced it
- Normalised part number parsing across manufacturer numbering conventions
- Dimensional and specification search for cases where no number is available at all
- Supersession chains, so a discontinued part resolves to its current replacement
- Explicit unknown responses rather than a low-confidence guess presented as a match
Endpoints
| Method | Path | Returns |
|---|---|---|
| GET | /v1/parts/{number}/equivalents | Equivalents for a part number with equivalence type |
| POST | /v1/parts/search | Find candidates from dimensions and specifications |
| GET | /v1/parts/{number} | Normalised part record with parsed attributes |
| POST | /v1/parts/batch | Resolve a whole bill of materials or stock list |
| GET | /v1/parts/{number}/supersessions | Supersession chain for a discontinued part |
| POST | /v1/feedback | Confirm or reject a match from a real fitment outcome |
Request and response
illustrativeRequest
{
"number": "SKF-6205-2RSH",
"context": { "equipment": "centrifugal pump, 2900 rpm", "duty": "continuous" },
"min_confidence": 0.6,
"include": ["spec_comparison", "supersessions", "stocked_only"]
}Response
{
"query": { "raw": "SKF-6205-2RSH", "parsed": { "manufacturer": "SKF",
"family": "6205", "suffix": "2RSH" } },
"resolved_part": {
"type": "deep_groove_ball_bearing",
"bore_mm": 25, "outer_mm": 52, "width_mm": 15,
"seal": "contact_rubber_both_sides", "dynamic_load_c_kn": 14.0,
"limiting_speed_rpm": 8000
},
"equivalents": [
{
"number": "NSK-6205DDU",
"manufacturer": "NSK",
"equivalence": "direct_replacement",
"confidence": 0.96,
"differs_in": [],
"evidence": ["dimensional_identical", "seal_type_identical",
"published_interchange_table"]
},
{
"number": "FAG-6205-2RSR",
"manufacturer": "FAG",
"equivalence": "functional_equivalent",
"confidence": 0.89,
"differs_in": [
{ "attribute": "dynamic_load_c_kn", "query": 14.0, "candidate": 14.3 }
],
"evidence": ["dimensional_identical", "seal_type_equivalent"]
},
{
"number": "GEN-6205-2RS",
"manufacturer": "Generic",
"equivalence": "dimensional_match",
"confidence": 0.62,
"differs_in": [
{ "attribute": "limiting_speed_rpm", "query": 8000, "candidate": 6300 },
{ "attribute": "temperature_max_c", "query": 120, "candidate": 80 }
],
"caution": "Lower speed and temperature rating; verify against duty"
}
],
"unresolved": [],
"note": "Fitment suitability remains an engineering judgement; specification differences are reported, not adjudicated."
}Where the data comes from
| Manufacturer catalogues and datasheets | Published specifications, used within each publisher's terms — several permit lookup but not bulk redistribution |
| Published interchange tables | Where manufacturers publish their own cross-references, which are the strongest evidence available |
| Your own catalogue and stock data | Your part records, descriptions and inventory, which is what makes results commercially useful rather than academic |
| Standards reference data | ISO, DIN and ANSI dimensional standards, which give many part families a common specification frame |
| Your historical substitution records | Past orders where one part was supplied against another number — real-world equivalence evidence and usually untapped |
| Distributor catalogue data | Where licensing permits; terms vary widely and bulk ingestion is often explicitly prohibited |
The intelligence layer
The difference between a demo and a service is mostly in this section. These are the parts that keep the answer trustworthy once real traffic is hitting it.
- Part number parsing per manufacturer convention, because the suffix is usually where the meaningful specification sits and a generic tokeniser destroys it
- Specification extraction from datasheets with unit normalisation and tolerance handling, since the same attribute is published in different units and conventions
- Attribute-level matching rather than whole-record similarity, so the response can state exactly which specifications differ
- Graded equivalence classes with explicit definitions, because collapsing dimensional match and direct replacement into one flag is the error that gets equipment damaged
- Confidence from evidence strength — a published interchange table outranks an inferred dimensional match, and the response says which it was
- Supersession chain resolution across multiple generations, which matters most for exactly the old equipment people are trying to keep running
- Validation against your historical substitution records, which is the only ground truth that reflects what actually fitted
- An explicit no-match response with the reason, since a confident wrong match costs far more than an honest blank
What it changes for the business
Searches that convert
An unstocked number returns stocked equivalents instead of nothing, which turns a lost search into an order.
Counter time reduced
Staff stop phoning round to establish equivalence that the system can state with evidence.
Knowledge out of one person's head
Informal cross-reference expertise becomes a maintained dataset rather than a retirement risk.
Obsolescence handled
Supersession chains keep old equipment serviceable instead of turning every discontinued part into a research task.
How it fits into your product
How it embeds
- Search endpoint callable directly from your catalogue search box, with a latency budget that suits typeahead
- Batch resolution for bills of materials, stock files and customer part lists
- Equivalence results annotated with your own stock position and pricing
- Feedback endpoint so confirmed and rejected fitments improve matching over time
- Deployment inside your infrastructure where catalogue and customer data cannot leave
- Per-tenant catalogue scoping for marketplaces and multi-brand distributors
Included in every build
- Authentication, scoped API keys and per-customer rate limiting
- Input validation with predictable, documented error responses
- Versioned endpoints so your integration never breaks without notice
- Request and response logging for audit and dispute resolution
- Latency, error-rate and usage monitoring with alerting
- OpenAPI specification, sandbox environment and client examples
- Backups, failover design and a documented recovery procedure
- An agreed support arrangement once the service is live
How we would build it
about 12 weeksDiscovery and data review
Weeks 1–2
We agree the exact question the API answers, review the sources you already hold, check what has to be licensed, and write down the response contract your developers will code against.
Architecture and prototype
Weeks 3–4
You get a written architecture plus a working endpoint on real data — narrow, but genuinely callable — so integration can be tested against something concrete rather than a document.
Production build
Weeks 5–10
Pipelines, models, storage, authentication, rate limiting, logging, tests and documentation. Security and observability are part of this phase, not a later one.
Integration and handover
Weeks 11–12
We support your engineers through their integration, run load tests at your expected volume, and agree how the service is monitored and supported once it is live.
Commercial models
There is no fixed price list, because the cost is driven by data complexity, integration count, model requirements, request volume and the support level you need. Scope is agreed in writing before development starts.
Fixed-scope prototype
A defined slice of the service, built and priced as one piece of work. The usual first step when the business case needs proving internally.
Fixed-price build
An agreed scope against a written architecture, delivered for an agreed price. Best when the requirements are already clear.
Build then retainer
We build the service, then operate and improve it under a monthly agreement covering monitoring, updates and model maintenance.
White-label product
The service runs under your brand, your domain and your pricing, with MLPivot invisible to your customers.
Questions buyers ask
Can you guarantee a match will fit?
No, and we would not build a product that implies it. The API reports dimensional and specification equivalence with the differences named. Whether a part is suitable depends on the application — duty cycle, temperature, environment, approvals, what the equipment manufacturer's warranty says. That judgement stays with the engineer, and the response is written to support it rather than replace it.
What about safety-critical and certified parts?
Those need handling differently and we would scope them out of automatic equivalence. Pressure-rated, ATEX, aerospace, lifting and other certified components have approval requirements that dimensional matching cannot address. For those families the API returns candidates flagged as requiring certification verification, and we would recommend your interface treats them as research leads rather than substitutions.
Can we ingest competitors' catalogues?
Usually not in bulk. Manufacturer and distributor catalogue terms frequently permit individual lookup while prohibiting systematic collection or redistribution, and some are explicit about it. We check the terms of every intended source during discovery and design the coverage around what is permitted. Where a source cannot be used, your own catalogue plus standards data plus published interchange tables usually still supports a worthwhile build — just a narrower one than a scrape would give, and we would rather say that up front.
How good is matching on parts with no published specifications?
Weak, and the API says so instead of guessing. Unbranded, obsolete and locally-made parts often have no datasheet anywhere. Dimensional search helps when someone can measure the item. Otherwise the honest answer is an unresolved response with the attributes we could not establish.
How much does our own order history help?
More than anything else in the build. Records of one part supplied against another number, with no return or complaint following, are direct evidence of real-world equivalence, and no published source matches that. It is also usually the least-examined dataset a distributor owns.
Related blueprints
Industrial Supplier Intelligence API
Capacity, financial distress and disruption signals for the suppliers a manufacturer depends on, delivered as monitoring rather than a one-off report.
Equipment Resale Valuation API
Used-equipment valuations and depreciation estimates for industrial machinery, served to lenders, dealers and asset managers.
Predictive Maintenance API
Failure probabilities, anomaly scores and maintenance recommendations for industrial equipment, served to your CMMS or plant software.
Want this built for your platform?
Tell us what your product does today and what you want another system to be able to ask it. We will come back with an approach, a scope and a price.