Customer Stories / Financial Services / Brazil

Empowering Employees to Work Strategically Using Amazon Bedrock with BDM
Learn how Amazon Bedrock helped BDM transform document processing.
85%
faster analyses
98%
accuracy
40%
reduction in operational effort
50%
reduction in development costs
Accelerated
decision-making
Overview
Manually analyzing complex documents is a painstaking but familiar process for many industries, from financial services to the legal sector and beyond. It takes a huge amount of time, resources, and effort. Every day employees spend hours sifting through large volumes of unstructured data. And by the time they have extracted and interpreted that data, human error can prevent accurate analyses.
Big Data Mining (BDM) set out to solve this problem and make people’s day-to-day work more rewarding with the help of artificial intelligence (AI). Founded in Brazil back in 2018, BDM began by tackling a scenario that frustrates many consumers—searching for the best car to suit their needs within a crowded marketplace. They dramatically simplified customers’ buying experience by automating the process and transforming masses of data into valuable information.
Fast-forward to today, and BDM has created LOUIS alongside AWS—a bespoke generative AI model that transforms document processing across a wealth of industries with over 100 models. The company is now enabling actionable insights from a vast range of different contexts to help people and organizations do more with less.

Opportunity | Fine-tuning AI with Amazon Bedrock for 85% faster analyses
Document processing can be a momentous task where there are millions of documents to analyze. The process is even more complicated when grappling with unstructured data from a variety of formats, such as images, videos, and audio files. Documents can also be subject to several updates during the processing period, making the pace of review critical to accuracy. This is especially important when dealing with sensitive information, such as contracts and powers of attorney, where outdated information could have serious consequences.
LOUIS solves these workflow issues and empowers companies to be more agile. From its inception, BDM built LOUIS on Amazon Bedrock, where they could access and customize high-performance foundation models from leading AI companies. They used the service to combine computer vision and natural language processing technologies to handle a range of complex corporate documents at scale. Not only does this make the process 85 percent faster than manual interpretation, but customers also achieve 98 percent accuracy. All of this means that businesses can spend more time on high-value, strategic tasks to help them gain a competitive advantage.

LOUIS cut operational effort by 40 percent—saving $4.2 million over five years.”
Deric Ferreira
Founder, Big Data Mining
Solution | Reducing operational effort by 40% with generative AI
A large financial institution in Brazil is just one customer benefitting from greater efficiency and scale by deploying LOUIS. Before the implementation, every month a team of 150 professionals needed to carefully evaluate over 40,000 unstructured and standardized processes in order to open legal entity accounts to maintain and contract credit. The process was prone to errors and decision-making often relied on other employees to offer a second opinion and re-analyze the documents.
LOUIS dramatically sped up the workflow by creating a process with standardized criteria for interpreting powers. This reduced operational effort by more than 40 percent—the equivalent cost saving of over $4.2 million over five years. AI has also saved employees from repetitive tasks, freeing them to focus on delivering human value, and consultative expertise. Now they are using their skills to analyze complex stock chain operations and strategically assess risks, rather than just transcribing and re-typing information into various systems.
The success of the project enabled BDM to expand the service to other financial institutions and explore new ways to extend human capabilities with generative AI.
Outcome | Building a foundation for experimentation with AWS services
Since then, LOUIS has been applied to 15 different industries—now over 20 proofs of concepts are creating powerful results. In the aviation space, LOUIS can assess various geographic regulations, taxes, and calculations to recommend the most cost-effective locations to refuel aircrafts. Another model is reshaping end customers’ experiences within insurance through rapid response times. Before using the solution, life insurance claims took up to six months to process due a complex queue of demands across insurers’ different business areas. LOUIS has accelerated decision-making down to just a few minutes by capturing the data required in just 60 seconds.
By using a range of services such as Amazon S3, Amazon EC2, and Amazon SageMaker, BDM has opened the door to numerous future applications. They can realize secure, scalable, and profitable outcomes through AWS’s commercial support, architecture reviews, and introductions to potential prospects. The pay-as-you-go-model for Anazon Bedrock has proven invaluable for helping them take advantage of the huge computing power needed to develop iteratively, while keeping costs under control. The organization has saved approximately 50 percent on development costs by using Amazon Bedrock versus writing the application from scratch.
With flexible and robust infrastructure, BDM can innovate without common time constraints. Tailoring pre-trained models to industry-specific needs significantly reduced their time to market and helped them outperform other advanced large language models in analysis efficiency. As they continue to lead the development of solutions that transform people’s daily work, BDM is excited to discover new possibilities and technologies that make customers’ lives easier.
About BDM
BDM uses artificial intelligence to read, understand, and analyze data, regardless of its format. They empower people to focus on what really requires human intelligence, such as making strategic decisions or creating new things.
AWS Services Used
Amazon Bedrock
Amazon Bedrock is a fully managed service that offers a choice of high-performing foundation models (FMs) from leading AI companies like AI21 Labs, Anthropic, Cohere, Meta, Mistral AI, Stability AI, and Amazon through a single API, along with a broad set of capabilities you need to build generative AI applications with security, privacy, and responsible AI.
Learn more »
Amazon SageMaker
Amazon SageMaker is built on Amazon’s two decades of experience developing real-world ML applications, including product recommendations, personalization, intelligent shopping, robotics, and voice-assisted devices.
Learn more »
Amazon S3
Amazon Simple Storage Service (Amazon S3) is an object storage service offering industry-leading scalability, data availability, security, and performance.
Amazon EC2
Amazon Elastic Compute Cloud (Amazon EC2) is a web service that provides secure, resizable compute capacity in the cloud. Access reliable, scalable infrastructure on demand. Scale capacity within minutes with SLA commitment of 99.99% availability.
Learn more »
Get Started
Organizations of all sizes across all industries are transforming their businesses and delivering on their missions every day using AWS. Contact our experts and start your own AWS journey today.