Product Insight

Streamlining Document Submission with the help of AI

Nov 28, 2024
3 min read
In Document Management, ensuring accurate and consistent document submissions is essential. With the help of ChatGPT and Forms Regognizer a quality check of the documents can be performed automatically.
Johnny Vik
Johnny Vik

Introduction

In Document managment, ensuring accurate and consistent document submissions is essential. Contractors frequently submit documents to projects or clients, and it is common practice for a document controller to manually verify that key metadata—such as document numbers, titles, revisions, and discipline codes—match the document plan uploaded earlier. This process, while crucial, can be time-consuming and prone to human error.
To address this challenge, Omega 365 leverages the power of artificial intelligence to automate and enhance document checks, ensuring both accuracy and efficiency.

Automated Document Verification Workflow

  1. Contractor Uploads the Document
    When a contractor uploads a document, it enters the system for automated analysis. This step marks the beginning of the document verification process.
  2. Extracting Text with Form Recognizer
    To analyze the document, we use Microsoft Form Recognizer to extract text from its first page. Form Recognizer is particularly effective because it extracts text in a logical, chronological order, making it easier for AI to interpret the data accurately.
  3. Analyzing the Document Using ChatGPT
    Once the text is extracted, we use ChatGPT to analyze the document metadata and identify potential mismatches. The following prompt is used to guide the analysis:
    You are an intelligent assistant helping with document analysis. 
You will need to analyze the document and provide the following information: 
- Extract relevant metadata: DocumentID, Title, Revision, RevisionDate, Project, Parcel, ProducedBy, CreatedBy, VerifiedBy, ApprovedBy.
- Check for document number mismatches and note any inconsistencies.
- Check for revision number mismatches and note any inconsistencies.
Return the results in a JSON string format suitable for parsing into a C# object.
    The AI returns a structured JSON object containing:
    • Extracted metadata.
    • Flags for mismatches in document or revision numbers.
    • Comments detailing observed inconsistencies, if any.
  4. Cross-Referencing with the Document Register
    The extracted metadata is then compared against the document register, a pre-approved database of document plans. Any discrepancies flagged by the AI are highlighted for further review.
  5. Flagging and Distribution
    • If discrepancies are found, they are flagged for manual resolution.
    • If the document passes the checks, it is either distributed automatically for review or made available for informational purposes.

Example of extraction of data from the front page.

Benefits of AI-Driven Document Analysis

Integrating AI into the document verification process offers several advantages:
  • Efficiency: Automating the initial verification steps saves significant time and effort for document controllers.
  • Accuracy: AI minimizes the likelihood of oversight by meticulously analyzing metadata and identifying inconsistencies.
  • Scalability: With AI handling large volumes of documents, projects can scale their operations without a proportional increase in administrative workload.
  • Streamlined Workflow: Approved documents can be seamlessly distributed, ensuring faster collaboration and project progress.

Enhancing Collaboration with Omega 365

By combining AI tools like Form Recognizer and ChatGPT with Omega 365’s robust platform, we have created a solution that bridges the gap between manual verification and automated intelligence. This approach ensures that contractors’ documents are reviewed efficiently and accurately before reaching clients, enabling smoother project workflows and fostering greater trust between all stakeholders.