Module: Development of the Intelligent Procurement Management System (IPMS)

Introduction
The development of the Intelligent Procurement Management System (IPMS) follows
a structured Systems Development Life Cycle (SDLC) framework, ensuring a methodical
approach from conception to deployment. This module details the preparation stage, and
system design (including architecture, user interface mockups, and database schema). By
adhering to this model, the project guarantees a robust, scalable, and efficient system that
transforms procurement into a strategic function.

1. Follow a development model

Our project will follow the Agile methodology, specifically using the Scrum
framework. This means we will build the system in iterative, time-boxed cycles
called sprints (typically 2 weeks each), allowing for continuous feedback and adaptation.
The development process is broken down into stages. The first and most critical stage
is Preparation.

Key activities in the preparation stage include:

1.1 Knowledge Base: Articles and Books
These resources provide the theoretical foundation and industry best practices for developing
the IPMS:
 Procurement & Digital Transformation:
o Reports/Articles: Digital procurement: The new frontier for value
creation (McKinsey & Company, 2023) and AI-augmented procurement:
Transforming the source-to-pay journey (Deloitte Insights, 2022) to
understand strategic value creation and AI application in procurement.
o Academic Research: "The Impact of Digital Transformation on Procurement
Performance" (Handfield, R., et al., 2022) for evidence-based insights.
 Technology & Standards:
o Frameworks & Documentation: Official documentation for Django REST
Framework, [Link], and Scikit-learn to guide technical implementation.
o Industry Standards: *IEEE Std 1012-2020 - Standard for System and
Software Verification and Validation* to ensure software quality and
reliability.
 Scientific and Technical Databases: Access to online libraries (e.g., IEEE Xplore,
ACM Digital Library) for up-to-date research on AI/ML models, system architecture,
and API design.
1.2 Existing Related Software (Competitor Analysis & Benchmarking)
Analysis of established commercial systems helps in defining the feature set, user experience,
and technical approach for the IPMS.
  Enterprise Solutions: Analysis of modules within comprehensive ERP systems
like Oracle NetSuite and SAP Ariba for their supplier management and procurement
automation capabilities.
 SME-Focused Solutions: Review of platforms like Procurify and Spendwise to
understand usability and feature expectations for small to medium-sized businesses.
1.3 Stakeholder and Feasibility Analysis
 Identifying Stakeholders: Procurement officers, department managers (as
requisitioners), finance team, suppliers, and senior management.
 Defining Goals: Reduce procurement cycle time, achieve cost savings through spend
analysis, improve supplier performance, and ensure policy compliance.
 Assessing Feasibility: Budget for AI/ML tools, data availability and quality,
integration complexity with existing ERP/accounting systems, and organizational
readiness for process automation.
1.4 Software Tools (Development Stack)
The development requires a robust set of tools across the backend, frontend, database, and
specialized AI/ML fields.
 Backend: Python (Django/FastAPI) for core application logic and high-
performance, asynchronous APIs, leveraging its strong AI/ML ecosystem.
 Frontend: [Link] or [Link] for building dynamic, manager-friendly dashboards and
user interfaces.
 Database: PostgreSQL for reliable, transactional storage of POs, suppliers, and
spending data.
 AI/ML & Analytics: Python libraries (scikit-learn, Pandas, NumPy) for spend
classification, predictive analytics, and data processing.
 Document Processing: TensorFlow/PyTorch for developing intelligent OCR to read
invoices and contracts.
 Cloud & DevOps: Docker for containerization, Kubernetes for
orchestration, AWS/Azure cloud platforms, and GitHub Actions for CI/CD
automation.
Deliverables: Project charter, detailed requirement specifications, technology stack
document, and a high-level project plan.

2. System Design
2.1 System Architecture
The IPMS It determines the high-level organization of the system, showing how all the
parts interact.
 It would define how
o the User Interface (for procurement staff and managers),
o the Application Server (where the AI and data analytics run), and
o the Database communicate.
  It outlines the data flow (the "plumbing and electrical wires") for key processes like
submitting a purchase request, checking budgets, and processing an invoice.
 It specifies how external systems (like the accounting software) will connect to the
IPMS using APIs.
2.2 Interface Mockups (Wireframes)
These are simple drawings (wireframes) that show what the user will see, focusing on the
user experience.
 They would design the layout for the visual dashboard showing company spending
by category, department, and supplier.
 They would mock up the screens for submitting new purchase requests, managing
supplier RFQs, and viewing supplier performance ratings.
 They ensure the interface is intuitive for the various users (buyers, approvers,
managers) who rely on the useful intelligence for decision-making.
2.3 Database Structure
It designs how all the necessary information is stored, organized, and retrieved quickly.
 It creates the structure for the single, searchable database required to house all
supplier information, including contracts, performance ratings, and risk scores.
 It organizes data fields for purchase requests (item, cost, department) to ensure they
can be automatically routed for approval.
 It defines the tables and relationships for storing spending data to allow easy filtering
for the cost-saving analysis dashboard.
 This structure is crucial because the effectiveness of the smart data analysis relies
entirely on having good, clean data.
Table Name Description IPMS Objective Supported
Create a single, searchable database for all
Stores supplier information, including performance and
suppliers supplier information, including contracts,
risk data. (Already in Snippet)
performance ratings, and risk scores.
department Automatically route purchase requests
Stores all company departments.
s based on the department.
Stores all users (employees, managers, finance) who Route requests to the right manager for
users
interact with the system. approval.
Stores the expense categories for goods/services (e.g., Show company spending by category to
categories
IT, Marketing, Office Supplies). find cost-saving opportunities.

Table Name Description IPMS Objective Supported

Develop a system where purchase requests are
purchase_request The initial request from an employee before
automatically sent to the right manager for
s approval and conversion to a PO.
approval based on the item, cost, and department.
Stores the logic (e.g., IF cost > $500 AND Ensure purchase requests are automatically sent
approval_rules
department = 'IT', THEN approver = 'CTO'). to the right manager.
purchase_orders The legally binding document sent to the Automate source-to-pay procedures.
 supplier. (Already in Snippet)
Connect the procurement system to the
invoices Records invoices received from suppliers. accounting software to automatically check
budgets and process invoices.

PROTOTYPE:
3. Coding and Development
The coding/development stage translates the IPMS design into a working prototype, focusing
on core automation and intelligence features. Development follows a modular, service-based
approach using Python and FastAPI for the backend logic and APIs.

3.1 Backend Development (APIs)

RESTful APIs are built to handle key CRUD (Create, Read, Update, Delete) operations for
the core procurement objects (requisitions, purchase orders, suppliers, and invoices).
 Technology: FastAPI (Python) is used for building high-performance, asynchronous
APIs.
 Key Operations:
o Managing supplier data (contracts, risk, and performance scores).
o Handling the submission and status updates of purchase requests
(requisitions).
o Creating and managing Purchase Orders (POs) and associated invoice records.
Sample API Endpoint (FastAPI/Python):
This endpoint provides useful intelligence by calculating a real-time supplier score, directly
supporting strategic sourcing decisions.
3.2 Automated Approval Routing Module
This module fulfills the objective to automatically send purchase requests to the right
manager for approval based on item, cost, and department.
 Logic: Uses a set of configurable business rules (e.g., if Category = 'IT' AND Cost >
$5,000, route to CTO) to determine the next approver ID.
 Integration: This logic is implemented as a microservice (approval_router.py) that is
called immediately after a new purchase request is created via the main API.

3.3 Spending Analysis & Reporting Module

This module processes large volumes of transaction data to create the summary metrics
needed for the visual dashboard.
 Function: Cleans, aggregates, and rolls up spending data by supplier, category, and
department.
 Output: Generates the data structures required by the frontend to display charts and
KPI cards, enabling staff to easily find cost-saving opportunities.

3.4 External System Integration (APIs)

To connect the procurement system to the accounting software to automatically check
budgets and process invoices.
 Methodology: Uses lightweight REST APIs to communicate with the company's
existing Enterprise Resource Planning (ERP) or accounting system.
 Operations:
o Budget Pre-Check: Querying the ERP budget ledger before final PO approval.
o Invoice Processing: Sending finalized invoice data to the accounting system
for payment (Automating source-to-pay).

3.5 Deployment (Blueprint)

While this project is limited to a design and prototype, the production deployment would
follow modern DevOps practices:
 Containerization: All Python services (API, Router, Analytics) would be
containerized using Docker to ensure consistency across environments.
 Continuous Integration/Continuous Deployment (CI/CD): Integration with
GitHub Actions or similar tools to automate testing and build processes.
 Cloud Infrastructure: Deployment to cloud platforms like Kubernetes or AWS
ECS, enabling high availability and scalability for large businesses.
REFERENCES:
 Intelligent Procurement Management System Project Document. IPMS-Project-
Specification-v1.0.
 Django Software Foundation. (2024). Django REST Framework: Web APIs for
Django. [Link]
 [Link] Documentation. (2024). React: A JavaScript Library for Building User
Interfaces. [Link]
 Scikit-learn Developers. (2024). scikit-learn: Machine Learning in
Python. [Link]
 McKinsey & Company. (2023). Digital procurement: The new frontier for value
creation. McKinsey Operations Practice.
 Deloitte. (2022). AI-augmented procurement: Transforming the source-to-pay
journey. Deloitte Insights.
 Handfield, R., et al. (2022). "The Impact of Digital Transformation on Procurement
Performance." Journal of Supply Chain Management, 58(3), 45-67.
 IEEE Standards Association. (2020). *IEEE Std 1012-2020 - Standard for System and
Software Verification and Validation*.