Use Case Intelligence RSS Search use cases Search
OpenAI Manufacturing Asia-Pacific & Oceania

Kuken Industries

Document Processing · Enterprise Knowledge Search

Kuken's engineers needed to find the correct replacement-part drawing in a 70-year archive. Even after digitization, searches depended on knowing historic part names and numbers, consuming experienced engineers' time and limiting newer staff's independence.

Read the original story Explore all use cases
4.5 minutes to find a drawing, down from nearly an hour
About 90% of issues flagged by the drawing-review prototype were genuine discrepancies
Projected 90% reduction in hours spent searching for drawings each year

Reported impact

Business outcome

The search tool reduces drawing retrieval from nearly an hour to 4.5 minutes and is expected to cut annual search hours by about 90%. Newer engineers can handle a wider range of work independently, while experienced engineers spend more time mentoring.

  • Productivity Improvement
  • Speed to Market

Solution

Kuken and Japanese AI startup Recursive built a drawing-search system and field app using OpenAI models through the OpenAI API. Engineers can enter a part or drawing number, or upload an image or PDF of a drawing or parts list. The app reads part numbers from the image or document and returns possible matches from Kuken's digitized archive, including drawings for similar designs. Engineers then compare the matched drawing with equipment specifications and replacement records to select the correct version for a customer installation. The system gives less experienced engineers a self-service way to find and study historic designs, while experienced staff gain more time for mentoring and engineering work. The source does not identify a specific model, archive-storage product, or retrieval infrastructure.

Data flow

An engineer enters a part number or uploads a drawing or parts-list image/PDF; the app extracts part numbers, searches Kuken's digitized archive, and returns candidate drawings for the engineer to verify against installation specifications.

Solution architecture

2 components · 2 layers
  1. Compute
    • OpenAI API Provides models that read drawing and parts-list details and support search within the engineering archive.
  2. Serving
    • Kuken drawing-search app Accepts part numbers, images, and PDFs and returns candidate historic drawings to field engineers.

Architecture clues

  • A separate drawing-review tool remains in development and is not included as a deployed capability
  • Engineers can search by part/drawing number or upload an image or PDF
  • OpenAI models were compared for extraction accuracy and cost

Evidence from the source

"Engineers can enter a drawing or part number, or upload an image or PDF of a drawing or parts list."
"Kuken expects the tool to cut the hours spent searching for drawings by about 90% over a year."
Transparency Evidence & methodology 100/100 reference value · 85% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value100/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact25/25
    3 metrics describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence85%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability25/25
    12 of 12 key fields link to a named source.
  • Direct evidence10/20
    2 direct source quotes retained.
  • Date provenance10/10
    The publication date comes from the source page or publisher.
  • Source transparency10/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. OpenAI

    Kuken Industries searches 70 years of drawings in minutes

    First-party
    https://openai.com/index/kuken/
    Source type
    Vendor-published customer story
    Published
    Nov 9, 2026
    Supports
    • Customer
    • Industry
    • Region
    • Company size
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • Business outcome
    • Metrics
    • Architecture
    • Publication date

Date evidence

  • Published on source page Nov 9, 2026 · OpenAI · First-party
    Open date source

Related

Similar use cases

See all like this
OpenAI Oct 28, 2025 Dai Nippon Printing Manufacturing Asia-Pacific & Oceania DNP needed to preserve institutional knowledge, standardize inconsistent expert work, and automate research, security, and operations tasks across a very large industrial enterprise facing labor constraints and complex… 100% weekly active usage rate Document ProcessingEnterprise Knowledge Search ChatGPT Enterprise · custom GPTs · internal knowledge base OpenAI Sep 24, 2025 ENEOS Materials Manufacturing Asia-Pacific & Oceania ENEOS Materials needed secure AI that could help a shrinking workforce perform research, plant design, training analysis, and other knowledge-heavy manufacturing tasks faster without exposing proprietary information. 80% of employees reported significant workflow improvements during pilot phase Document ProcessingEnterprise Knowledge Search ChatGPT Enterprise · custom GPTs · deep research OpenAI Nov 12, 2026 Gilbert + Tobin Commercial and Professional Services Australia Gilbert + Tobin wanted to improve client service and employee productivity while enabling a law firm to use AI safely with sensitive work and confidential client information. 87% active usage among enabled ChatGPT users Document ProcessingEnterprise Knowledge Search ChatGPT Enterprise · ChatGPT Work · Codex