aPriori APIs

Cost+ intelligence, at the point of decision

Automate should cost & DFM directly from PLM, and easily bring all of your data into BI to track value and progress.

Book a Demo

AUTOMATE CAD WORKFLOWS
No more manually downloading/uploading CAD files

AUTOMATE REPORTING
No more reworking spreadsheets just to build the leadership report

INTEGRATE INTELLIGENCE
Reduce the need to manually log in to aPriori applications

Automate with aPriori APIs

See aPriori's deep, rich cost & DFM insights directly where you work (PLM, CAD, and more)

Connected manufacturing intelligence

aPriori Cloud Platform APIs

Connect your applications to aPriori to run cost analysis and export data into your analytics tools.

Customer
Applications

CAD
PLM
ERP
AI agents
Submit analysis Send results back
aPriori Cloud
Role-Based Applications
aP Pro
aiDesign1
aiSource
aP Workspace
Intelligence API & MCP1 Submit analysis
Receive results
aPriori
Engine
Cost · DFM · Sustainability
Data Export API Data out
Analysis out
Data & Analytics

Customer
Analytics

Data Lake
BI
AI initiatives
→ Data out
Analysis out

1 MCP and aiDesign in development

What is the Data Export API?

A secure, standards-based interface that makes your aPriori analysis data available outside the aPriori - on your schedule, in your systems.​
  • REST API – works with any language or platform
  • OAuth 2.0 – enterprise-grade security
  • Flat JSONL & Parquet output loads into modern databases
  • Incremental delta exports – only pull what changed
  • Export data on scenarios, processes, assembly structure and users, with more data products on the way.
  • aPriori Cost analyses & DFM data
  • Data Export API Secure REST endpoint
  • Your Warehouse, Data Lake, BI Platform Upserted, always current
Your Application aPriori API RESULTS RETURNED Cost DFM Sustainability Inputs Used Status and Errors

What is the Intelligence API?

A secure, standards-based interface for programmatically submitting CAD files to aPriori for analysis.
  • REST API – works with any language or platform
  • OAuth 2.0 – enterprise-grade security
  • Submit up to 250 CAD files
  • Asynchronous – submit job and poll for status and results
  • Returns cost, DFM, and sustainability (if licensed) results
Top 5 API Use Cases

The cost+ answer, before the design is locked

Five teams at Corvane Industries, a fictional maker of compact loaders and hydraulic components. Each one gets cost+ insight inside the tool they already use, because aPriori's APIs connect it to their PLM, data warehouse, sourcing, quoting and planning systems.

THE OLD WAY: WITHOUT APRIORI
  1. Export the part files from PLM.
  2. Upload it and email the cost team.
  3. Wait. Follow up. Wait again.
  4. The design is already locked.
Autoplay

Use case 1 · Costing from PLM

Maya's bracket

Our engineers shouldn't have to export design files just to get a cost estimate.

EV manufacturer

Maya Okafor · Design engineer, CL-450 Loader

Rev D, no cost yet

Tuesday · 2:25 PM

Before

Rev D of the loader arm bracket needs a cost check.

Maya has finished Rev D of the LA-2210 bracket assembly for the CL-450 loader and is about to check it in to PLM. Until now, getting a cost meant exporting design files from PLM, emailing the cost team and waiting about two weeks. By then the design was locked.

PLM · Check in

2:31 PM

When she checks in Rev D, PLM costs the whole assembly.

PLM is the system of record, so costing starts there. When the assembly is checked in, all four parts go to aPriori in one batch automatically. Anyone can also run an analysis on demand from any PLM part or assembly record.

POST /v1/files ×4 · 4 file IDs
POST /v1/input-profiles ×2 · sheet metal + forging
POST /v1/jobs 201 · SUBMITTED · 4 files
POST /v1/jobs
{
  "entries": [
    { "inputProfileIdentity": "Ip7Sm3Kc9Rt2",
      "fileIdentities": ["Fm1Pl8Xr4Kd2", "Fg2Us5Tq9Wb7", "Fs3Ch7Nl2Qw9"] },
    { "inputProfileIdentity": "Ip8Fg4Ld2Nm6",
      "fileIdentities": ["Fp4Bs6Jv1Hy3"] }
  ]
}

PLM · Batch analysis

2:31 – 2:34 PM

aPriori costs every part in the assembly.

aPriori builds a digital manufacturing model for each part: process routing, cycle time, material, labor, overhead and carbon. The gusset's model has no solid body, so only that part is flagged. The other three finish normally.

GET /v1/jobs/Jb8Rw2Lx5QnA PROCESSING · Retry-After 15
GET /v1/jobs/Jb8Rw2Lx5QnA COMPLETED
HTTP/1.1 200 OK
Retry-After: 15

{ "jobIdentity": "Jb8Rw2Lx5QnA", "status": "PROCESSING",
  "submittedAt": "2026-10-23T14:31:16Z", "updatedAt": "2026-10-23T14:31:31Z" }

PLM · Results on record

2:34 PM

Fixed while it's cheap

Results land on every part record, for the whole team.

Cost, DFM and carbon now sit on each PLM part record, where cost engineers, buyers and planners can see them. Rev C of the side channel came back at $24.63 with high DFM risk, so Maya opened up the bend radius, moved the hole away from the bend and relaxed the flange tolerance. Rev D shows $17.62, under target with low DFM risk. That's $84,120 a year saved before any tooling is cut. The gusset failed, so PLM opens a task for Maya to re-export it.

GET /v1/jobs/Jb8Rw2Lx5QnA/results 3 succeeded · 1 failed

Use case 2 · Cost data in BI

The report that builds itself

We want to stop reworking spreadsheets every month just to build the leadership report.

EV manufacturer

Sam Reyes · Cost analyst, Program Office

Month-end scramble

First Monday of the month

Before

Sam rebuilds the leadership report. Again.

Every month Sam exports scenario data, cleans it in spreadsheets and rebuilds the CL-450 cost report by hand. It takes two days, and by the time it reaches leadership the numbers are already out of date.

Connect once

Setup · one time

Connect aPriori to the warehouse once.

The data team picks Scenario Cost Summary from aPriori's data catalogue. It holds the latest fully burdened cost for every scenario, one row each. Its data contract defines every field, so the warehouse table is built straight from it.

GET /v1/data-products 3 data products
GET /v1/data-products/{id}/contract ODCS v3.1.0 · v2
{ "name": "scenario_cost_summary",
  "properties": [
    { "name": "scenario_id", "logicalType": "string", "primaryKey": true },
    { "name": "fully_burdened_cost", "logicalType": "number" },
    { "name": "meta_incremental_key", "logicalType": "date" }, … ] }

First load

Setup · first load

All of 2026 arrives in one pull.

The first export brings in every scenario since January as Parquet files. They're loaded into the warehouse and merged on the contract's primary key. It takes a few minutes, once.

POST /v1/data-products/{id}/exports incremental · parquet · RUNNING
GET /v1/data-products/{id}/exports/{exportId} SUCCEEDED · 48,213 records
GET …/exports/{exportId}/results 3 pre-signed file URLs
POST /v1/data-products/3f6c2a9e-…/exports
{ "dataExportType": "incremental", "format": "parquet",
  "dataStartAt": "2026-01-01T00:00:00Z",
  "dataEndAt":   "2026-10-23T06:00:00Z" }

Nightly changes

Every night · 06:00 UTC

After that, it pulls only what changed.

Each night the pipeline asks for records changed since the newest one already in the warehouse. Most nights that's a few hundred records: new parts, design changes such as Maya's Rev D, and rate updates.

SQL MAX(meta_incremental_key) 2026-10-23T05:58:41Z
POST /v1/data-products/{id}/exports from cursor · 312 records
-- start from the newest record you already have
SELECT MAX(meta_incremental_key) FROM analytics.apr_scenario_cost_summary;

POST /v1/data-products/3f6c2a9e-…/exports
{ "dataExportType": "incremental", "format": "parquet",
  "dataStartAt": "2026-10-23T05:58:41Z", "dataEndAt": "2026-10-24T06:00:00Z" }

Monday dashboard

Monday · 8:00 AM

No spreadsheets

On Monday morning, the dashboard is already current.

Leadership opens the program review and sees last night's numbers, including the $7.01 per loader saved by Maya's Rev D bracket. Sam spends those two days on analysis instead of rebuilding spreadsheets.

Use case 3 · Should-cost in sourcing

Lena's negotiation

Should-cost should already be in the sourcing tool when the RFQ goes out.

Home appliance maker

Lena Marsh · Commodity buyer, Castings

RFQ, no should-cost

Wednesday · 9:10 AM

Before

Lena is about to send an RFQ blind.

Lena is sourcing the CV-40217 hydraulic valve body: 25,000 a year for four years. A should-cost from the cost team takes about a week, so most RFQs go out without one and quotes get compared in Excel.

Should-cost attached

9:12 AM

The should-cost arrives before the RFQ goes out.

When Lena publishes the RFQ, the sourcing tool sends aPriori the same STEP model the suppliers get. It's costed as a die casting made in Mexico, where the suppliers are. Two minutes later the RFQ shows a $31.47 should-cost and what drives it.

POST /v1/files CV-40217_valve-body.stp
POST /v1/input-profiles Casting - Die · aPriori Mexico
POST /v1/jobs SUBMITTED
GET /v1/jobs/{id}/results 204 wait → 200 · $31.47
POST /v1/input-profiles
{ "processGroupName": "Casting - Die",
  "materialName": "Aluminum, Cast, ANSI AL380.0",
  "digitalFactoryName": "aPriori Mexico",
  "annualVolume": 25000, "productionLife": 4,
  "machiningMode": "MAY_BE_MACHINED",
  "userDefinedAttributes": { "RFQ": "RFQ-26-0418" } }

Quotes checked

The following week

Every quote is checked as it arrives.

Three suppliers respond. Each quote is compared with the should-cost as soon as it arrives, and any quote more than 15% above it is flagged. Lena can see at a glance which conversations matter.

Negotiate & award

Award

Savings tracked

She negotiates from the number, and the savings count.

Lena takes the cost breakdown into the call with Northgate Foundry. They agree on $32.40, down from $33.10 and within 3% of should-cost. The savings are logged automatically when she awards the business.

Use case 4 · Cost-based quoting

Dana's same-day quote

Quote faster and more accurately from a real cost model, inside your own quoting system, and win more business.

Dana Whitfield · Quoting engineer, Corvane Hydraulics

RFQ backlog

Monday · 8:30 AM

Before

Dana has 47 RFQs and one estimator's worth of time.

Corvane Hydraulics machines custom manifolds for other OEMs. Every quote is built by hand in a spreadsheet from the customer's drawing. It takes nine days on average, and by then some buyers have already gone with whoever answered first.

Estimate in one click

10:05 AM

The customer's CAD goes straight to aPriori.

Dana opens Halden Agri's RFQ for three manifold blocks and clicks Estimate. The quoting tool sends the models to aPriori with one profile per volume break the customer asked about: 500, 2,000 and 5,000 a year. That's nine analyses in one job.

POST /v1/files ×3 customer models
POST /v1/input-profiles ×3 · one per volume break
POST /v1/jobs 3 entries · 9 analyses
POST /v1/jobs
{
  "entries": [
    { "inputProfileIdentity": "Ip1Vb5Kq8Zt3",
      "fileIdentities": ["Fq1Hm3Bk7Tz2", "Fq2Hm5Ld8Wx4", "Fq3Ec6Rn2Py9"] },
    { "inputProfileIdentity": "Ip2Vb7Mx4Lc6",
      "fileIdentities": ["Fq1Hm3Bk7Tz2", "Fq2Hm5Ld8Wx4", "Fq3Ec6Rn2Py9"] },
    { "inputProfileIdentity": "Ip3Vb9Ws2Hd8",
      "fileIdentities": ["Fq1Hm3Bk7Tz2", "Fq2Hm5Ld8Wx4", "Fq3Ec6Rn2Py9"] }
  ]
}
# each profile: "Stock Machining", "Aluminum, 6061", annualVolume 500 / 2000 / 5000

Price every break

10:09 AM

Every volume break is priced from a real cost model.

Cost comes back for each part at each volume, built from aPriori's process model rather than a rule of thumb. Dana sets the margin and the prices update. aPriori also flags that the 4-port block's cross-drilled port needs a 5-axis setup. Try the margin slider.

GET /v1/jobs/{id}/results 9 succeeded · cost, DFM, cycle time
{ "componentName": "HM-312_manifold-4port", "status": "SUCCEEDED",
  "cost": { "fullyBurdenedCost": 101.60, "materialCost": 18.40, "currencyCode": "USD" },
  "dfm":  { "riskRating": "MEDIUM", "costDrivingIssuesCount": 1 },
  "cycleTime": 1284.0,
  "inputsUsed": { "annualVolume": 2000, "processGroupName": "Stock Machining" } }

Same-day quote

2:40 PM

Same day, not nine days

The quote goes out the same day, with a cheaper option.

Halden gets a quote covering all three volume breaks, plus an optional design change that takes 14% off the 4-port block. Offering a lower cost gives Halden a reason to choose Corvane. The lead time comes from the plant's real capacity plan.

Win more

Two quarters later

Illustrative

Faster, more accurate quotes win more business.

With quotes going out the same day, Corvane Hydraulics' win rate rises from 18% to 27%. Because prices start from a cost model, the actual cost of won work now lands within 4% of the quote.

POST /v1/data-products/{id}/exports nightly · Scenario Cost Summary
SQL JOIN erp.actual_cost quoted vs actual

Use case 5 · Capacity & lead-time planning

Owen's capacity plan

Bring aPriori routings and cycle times into ERP and MES to plan machine capacity and promise lead times.

Owen Brandt · Manufacturing planner, Plant 2

Old standards

Q4 planning review

Before

Owen is planning a ramp he can't see.

The CL-450 loader ramps to 12,000 units a year in Q1. Owen's capacity plan runs on routing standards from a 2019 time study, plus estimates for every new part. Nobody can say whether the press brakes will cope.

Routings in

Every night · 06:00 UTC

Routings and cycle times flow into ERP from aPriori.

Every part aPriori costs already has a modelled routing: operations, work centers, setup and cycle times. The Data Export API sends that process detail to ERP every night. New parts such as Maya's Rev D side channel get a routing as soon as they're costed.

POST /v1/data-products/{processDetailId}/exports incremental · parquet
GET …/exports/{exportId}/results 312 routings changed
SQL MERGE INTO erp.routing_staging on scenario + op
POST /v1/data-products/b2d9e4f1-6a73-4c08-8e5d-91f3a7c2b6d0/exports
{ "dataExportType": "incremental", "format": "parquet",
  "dataStartAt": "2026-10-23T05:58:41Z", "dataEndAt": "2026-10-24T06:00:00Z" }
# Manufacturing Process Detail: routings, process steps, cycle times, machine rates

Machine load

Same morning

The plan shows the bottleneck six months early.

Modelled cycle times multiplied by program volumes give the projected load on every work center. Press brake PB-01 passes 100% in March and peaks at 121% in June. Without this, the first sign would have been late orders.

Fix & promise

Planning review

Lead times you can keep

Test the fixes, then promise a lead time you can keep.

Owen compares three options. For the outsourcing option, the Intelligence API re-costs the part in aPriori's Mexico digital factory. Moving two of the four bends to the underused PB-02 costs just $0.12 a part and keeps every machine under 95%. Sales can now promise five weeks.

POST /v1/input-profiles digitalFactoryName: aPriori Mexico
POST /v1/jobs SUBMITTED
GET /v1/jobs/{id}/results $18.90 · cycle 161 s
POST /v1/input-profiles
{ "processGroupName": "Sheet Metal",
  "materialName": "Steel, Cold Worked, AISI 1020",
  "digitalFactoryName": "aPriori Mexico",
  "annualVolume": 12000, "batchSize": 500, "productionLife": 5,
  "scenarioName": "Outsource option B" }
Corvane Industries, its people, parts, suppliers and every number on this page are fictional and for illustration only. The screens are generic mock-ups, not any vendor's product.

Start sourcing designing making more profitable products today

See how aPriori helps engineering, procurement, and estimating teams accelerate development, reduce costs, and win more business with confidence.