Automating Procurement: Acme Manufacturing's Success Story with Claude-v2

Automating Procurement: Acme Manufacturing’s Success Story with Claude-v2

Business Problem

A global leader in industrial manufacturing, encountered significant inefficiencies in their vendor shortlisting and onboarding processes. The traditional manual approach required extensive human intervention at multiple stages, leading to high operational costs and lengthy processing times. The labor-intensive nature of these tasks not only slowed down the overall workflow but also increased the likelihood of errors and delays in vendor communications. These inefficiencies made it challenging to maintain accuracy and timeliness, which are critical for sustaining smooth operations and meeting production deadlines. As a result, Acme Manufacturing urgently needed a streamlined, automated solution to reduce costs, accelerate vendor selection and onboarding, and enhance the accuracy and reliability of their communications and documentation processes.

Solution

To address these challenges, GoML developed a multi-model Generative AI-powered solution using Claude-v2. The solution focused on automating key aspects of the vendor shortlisting and onboarding process:

Automated Vendor Shortlisting and Onboarding

Leveraged Claude-v2 to automate the vendor shortlisting and onboarding process, significantly reducing manual intervention and associated costs.

Vendor Documentation Processing

Utilized Claude-v2’s advanced text extraction and classification capabilities to automate the processing of vendor documentation, ensuring accuracy and efficiency.

Enhanced Communication

Implemented automated email communications to streamline interactions with vendors, ensuring timely and error-free correspondence.

Automating Procurement: Acme Manufacturing's Success Story with Claude-v2 — Click to View in Full Size

Architecture

Vendor Email Communication: Uses Amazon SES to send Requests for Proposals (RFPs) and registration forms to potential vendors, ensuring reliable and scalable email delivery.
RFP Upload and Analysis: Vendors upload RFPs, which are processed by AI models on AWS Bedrock, using advanced NLP and ML techniques for document analysis and classification.
Registration Form Upload and Processing: Vendors upload registration forms that undergo validation and processing to ensure all necessary information is accurately captured.
Document Processing and Validation: Extracted data from RFPs and registration forms is validated against organizational criteria and regulatory requirements using AI and relevance search algorithms.
Relevance Search: Matches vendor documents with the organization’s requirements, leveraging advanced search algorithms to ensure precise and relevant results.
Response Generation: Generates appropriate responses based on document analysis and validation, ensuring timely and accurate communication with vendors through Amazon SES.
FastAPI Orchestration: Manages API interactions and workflows efficiently, integrating various components and services within the architecture.