- Safety Engineering
Generative AI Solution for Efficient Knowledge Retrieval and Proposal Generation
JBS Dev built a GenAI tool on AWS for a safety engineering leader to retrieve internal docs and draft proposals.
Knowledge retrieval and proposals
Find internal information and prepare proposal drafts
AWS + generative AI
- Internal documents
- Knowledge search
- Proposal drafts
AT A GLANCE
The Project at a Glance
The Challenge
Staff at a safety engineering and risk mitigation firm reviewed large volumes of manuals, guidelines, proposals, and reports by hand to answer customers, analyze risk, and draft proposals.
What We Built
JBS Dev built a knowledge base and a proposal generation tool on AWS, using Amazon Bedrock, Claude Sonnet 3, AWS Lambda, Amazon OpenSearch, and Amazon RDS.
The Result
Employees find accurate information quickly instead of searching for hours. The proposal tool produces drafts that are 80 to 90% complete, so staff spend their time on customization and strategy.
BACKGROUND
Overview
Our client, a market leader in safety engineering and risk mitigation, specializes in forensic engineering, security risk consulting, and emergency management. To improve operational efficiency, they sought a generative AI (GenAI) solution deployed internally on AWS. The staff had to review a large volume of documents to respond efficiently, proactively address their customer needs, and perform a thorough risk and hazard analysis. This process was time-consuming and often required extensive manual effort, creating bottlenecks in their workflows. The goal was to help staff quickly access critical internal information from a vast repository of documents, including manuals, guidelines, proposals and reports. Partnering with JBS Dev, they embarked on a phased approach to develop and deploy this solution.
THE PROBLEM
Challenge
The client needed an AI-powered tool that could:
- Seamlessly integrate with their existing Single Sign-On (SSO) solution.
- Provide fast and accurate retrieval of internal documentation.
- Ensure reliability and precision in document processing while minimizing AI hallucinations.
- Offer a scalable framework to support ongoing improvements and expansions by their in-house team.
Given the sensitive and complex nature of the industry, the solution required advanced AI optimizations and secure deployment on AWS.
THE SOLUTION
Solution
JBS Dev partnered with the client to design and implement a scalable, high-performance GenAI solution leveraging AWS services. Our approach included:
INFRASTRUCTURE
1. Infrastructure & Optimization
- Designed an AWS architecture tailored for scale and performance.
- Implemented optimizations for AI model efficiency, including prompt engineering best practices and AI hallucination detection/prevention.
TECHNOLOGY
2. Core Technologies Used
- AWS Bedrock – Enabled the deployment and fine-tuning of the generative AI models.
- AWS Lambda – Provided serverless computing for document processing tasks.
- AWS OpenSearch – Allowed efficient indexing and searching of structured and unstructured data.
- AWS RDS – Managed the structured data repository.
- Claude Sonnet 3 – The LLM powering the solution, chosen for its balance of speed and accuracy.
COMPONENTS
3. Two Key Solution Components
- Knowledge Base Generation: Parsed and structured a vast library of manuals and guidelines into an easily searchable knowledge base, improving information retrieval for on-site personnel.
- Proposal Generation Tool: Built an AI-driven assistant to simplify and enhance proposal creation, reducing manual effort while ensuring accuracy and compliance.
Both components required extensive custom coding and sophisticated document processing techniques to ensure high-quality results.
THE OUTCOME
Results
Implementing the GenAI solution significantly improved how the client’s staff accessed and utilized internal documentation. Before this solution, employees faced significant challenges retrieving critical information due to the large volume of manuals, guidelines, and reports dispersed across multiple sources. Staff often spent hours manually searching through various repositories to find relevant materials. With the introduction of the AI-powered knowledge base, employees can now quickly retrieve accurate and relevant information and focus on high-value tasks rather than sifting through data.
The AI-driven proposal generation tool has also proven to be a game-changer for the client’s business operations. Previously, drafting proposals was a time-consuming and resource-intensive process requiring staff to gather information from disparate documents, often leading to inconsistencies and redundancies. Instead of creating each proposal from scratch, employees can now use the proposal generation tool to create a comprehensive, high-quality proposal that is 80-90% complete, ensuring consistency and maintaining company standards. This allows employees to focus on customization and strategy rather than spending excessive time on administrative tasks. Improved efficiency has also enabled them to respond faster to new opportunities, enhancing their competitive edge in the industry.
The internal team now has a powerful foundation upon which they can build further enhancements, ensuring long-term scalability and adaptability. By leveraging AWS services and implementing cutting-edge AI techniques, JBS Dev has delivered a transformative solution that has become an integral part of the client’s operational workflow.
KEEP READING
More Case Studies

Manufacturing
A GenAI Knowledge Base for a Building Materials Manufacturer
Employees ask questions in plain language and get answers almost instantly, from a knowledge base built on Amazon Bedrock and OpenSearch.
Read the building materials case study
Industrial Products
Agentic AI Knowledge Platform for an Industrial Products Provider
A RAG and agentic AI platform on AWS replaced it. Employees use it hundreds of times a week, and customers get source-backed answers about products.
Read the industrial products case study
Healthcare
Medical Record Extraction for a Healthcare Technology Company
Amazon Bedrock models and AWS Lambda now extract the data into a standard format, with less manual work and more accurate, timely reports.
Read the healthcare technology case study