Oil & Gas

Document Cleaning Automation

HomeCASESOil & Gas

[SOFTWARE DEVELOPMENT | NORWAY]

Overview:

Manual document cleaning was error-prone, with subjective interpretation leading to data inconsistencies.

Project in numbers

Industry: Construction and installation in O&G
Location: Bergen, Norway
Time period: 2024

Team size

Business analyst: 25% FTE
Senior Developer: 75% FTE

Tech stack

Python
Pandas
Openpyxl
Domain-specific rule modules

01CHALLENGE

High Error Susceptibility: The reliance on an entirely manual document-cleaning workflow significantly increased the likelihood of operational mistakes.
Labor-Intensive Execution: Requiring human intervention for every single cleaning and standardization step created a slow and inefficient process.
Output Inconsistencies: A lack of automated, uniform rules led to varying results, as multiple operators applied their own subjective interpretations to the data.

The customer had a manual document-cleaning workflow with an increased risk of errors. All cleaning and standardization steps were performed manually. Inconsistencies occurred due to subjective interpretation by multiple operators.

02SOLUTION

What we delivered

Improved data automation: Scripts auto-filled fields using predefined vocabularies stored in a local database.
Introduced domain-specific logic: Embedded rules handled edge cases and special scenarios using configurable Python logic.
Added SOP cheat sheet: A documented procedure outlined each step, ensuring consistency across operators.

We transformed the manual data-cleaning process into a fully automated, streamlined workflow. By integrating intelligent scripting and clear procedural guidelines, we ensured consistent data preparation and flawless application of complex domain rules.

03Results

Faster processing of document batches
Reduced manual error rates
Consistent application of business logic
Easier onboarding and process handover

Do you have a different challenge?

Get a free expert advice from us

Collaboration AI Assistant

Answers based on our experience and solved cases