Scribd
Upload
16 views4 pages

Parser Workflow

The document outlines a detailed workflow for an AI-based Construction Specification Parser and Checklist Generator aimed at automating the extraction and validation of construction project specifications according to Saudi Building Codes. It includes steps for user uploads, text processing, AI matching, checklist generation, and output storage, along with a user interface for viewing checklists and admin features for logging and management. The technology stack involves tools like React.js, Flask, and OpenAI's GPT-4 API, with future expansion ideas for OCR support and mobile applications.

Uploaded by

murtazamalik322
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
16 views4 pages

Parser Workflow

The document outlines a detailed workflow for an AI-based Construction Specification Parser and Checklist Generator aimed at automating the extraction and validation of construction project specifications according to Saudi Building Codes. It includes steps for user uploads, text processing, AI matching, checklist generation, and output storage, along with a user interface for viewing checklists and admin features for logging and management. The technology stack involves tools like React.js, Flask, and OpenAI's GPT-4 API, with future expansion ideas for OCR support and mobile applications.

Uploaded by

murtazamalik322
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

Here is a detailed, step-by-step workflow for your AI-based Construction

Specification Parser & Checklist Generator — designed to help your boss,


team, and developers clearly understand how the system works from start to
finish.

🧠 System Objective (1-liner):


Automatically extract and validate construction project specifications using Saudi
Building Codes (SBC), and generate tailored checklists for design review,
compliance, and approval.

🔁 End-to-End Workflow Overview


🔹 1. User Uploads Construction Spec
 📄 Input: Project spec (PDF, DOCX, or text) from client/engineer
 Stored At: uploads/ folder
 🔍 Backend Logic:
o Parse and extract full text from the uploaded spec

o Split into logical sections (e.g., concrete, HVAC, fire protection)

🔹 2. Text Processing & Pre-Matching


 📚 Codebook Lookup: Relevant Saudi Building Codes are stored in:
o shared_codebooks/

o Example: SBC-304_Concrete.pdf, SBC-201_Fire_Safety.pdf

 🧠 Extraction Tool: Use pdfplumber to extract text from both the spec and
relevant SBC documents.
 🤖 AI Matching:
o The system analyzes user text and identifies which sections of SBC
apply.
o Match phrases like: “compressive strength of concrete shall be...” →
trigger SBC-304, Section 6.2.

🔹 3. Prompt Generation & AI Response


 Custom Prompt Engine:
o Prompt created dynamically:

o You are a construction compliance checker for Saudi Arabia. Given the
following project spec and related SBC codes, generate a checklist for
reviewing concrete mix design compliance.
 🔗 OpenAI API Call:
o Send structured prompt with:

 Extracted spec section


 Relevant SBC clause
o AI responds with checklist (JSON or bullet format)

🔹 4. Checklist Generation & Output Storage


 📋 Checklist Types:
o Material Submittal

o Method Statement Review

o Compliance Review

o Design Validation

 💾 Saved To:
o /checklists/ folder

 JSON → for internal use


 Excel → for engineers/contractors
 PDF → for formal submissions
 📑 Sample Output File Names:
o checklist_concrete_section6_2025-07-19.json

o checklist_fire_systems_HVAC.pdf

🔹 5. Checklist Viewer (Web UI)


 🧑‍💻 User Interface:
o Dashboard with:

 Upload new spec


 View checklists
 Download reports
 Filter by project name/date/code type

🔹 6. Logging & Admin Features


 📝 Logs Saved in: /logs/
o checklist_history.csv

o user_uploads.csv

 🔐 Admin Options:
o View AI prompts/responses

o Compare AI checklist with manual checklists (for QA)

o Tag checklists as “approved”, “needs revision”, etc.

📲 Technology Stack

Layer Tool / Library

Frontend React.js + Tailwind + Axios

Backend Flask or FastAPI

AI Engine OpenAI GPT-4 API

PDF
pdfplumber or PyMuPDF
Parsing

Export xlsxwriter / pdfkit

Local (for now), S3 (future


Storage
scaling)

Logging CSV + JSON based logs

🧩 Modular Expansion Ideas (Post-MVP)


 🔍 OCR support for scanned specs
 📱 Mobile upload app for engineers
 📊 Analytics dashboard for code compliance frequency
 🔁 Integration with PM tools (e.g., Procore, Primavera)

You might also like