AI-Powered Intelligent Document Processing

Summary

CET, an Enterprise AI solution provider based in Singapore, addresses the inefficiencies of traditional OCR with their LM-based intelligent document processing. This solution is fast, cost-efficient, and supported by Nvidia GPUs, enabling financial service providers to automate credit analysis, fraud detection, and loan monitoring. With a unified framework supporting multiple models, CET has improved efficiency by over 50% and increased reasoning speed by 10%. Their Inception program and collaboration with Nvidia have been instrumental in expanding their market reach in APAC, positioning CET as a leading LM service provider.

Takeaways

• 🌟 CET is an Enterprise AI solution provider headquartered in Singapore.
• 🔍 Customers face issues with traditional OCR systems being slow, complex, and expensive.
• 📚 CET uses LM-based intelligent document processing for handling unstructured documents efficiently.
• 🚀 Support from Nvidia GPUs enhances the performance of CET's AI solutions.
• 📈 CET's FSA customers aim to grow their loan books without additional human resources, using CET's dashboard for automated data structuring and analysis.
• 🛠️ CET has transitioned from multiple backends to a unified framework, simplifying operations and improving efficiency by over 50%.
• 🔧 The reasoning speed of their AI models has increased by 10% due to the adoption of Fast Transformer and larger language models.
• 📱 CET's Inception program is highly valuable, particularly the APAC team's support in growing the business through various partnerships and events.
• 🔑 CET offers a solution called 'Intelligent Document Processing' to automate workflow from input to output.
• 🏦 Another solution, 'Fast Lender Workflow Automation', helps customers increase their loan books and make fast credit decisions without extra staff.
• 🤝 CET works closely with Nvidia and is poised to become one of the first LM service providers in the APAC region.

Q & A

What is CET and where is it based?

CET is an Enterprise AI solution provider based in Singapore.

What are the common issues customers face with traditional OCR according to the script?

Customers face issues such as traditional OCR being too slow, too complex, and too expensive.

What does CET use to process unstructured documents?

CET uses a Language Model (LM) based intelligent document processing to handle unstructured documents.

How does CET's solution benefit the FSA customers in increasing their loan book?

FSA customers can increase their loan book without additional human resources by uploading loan documents to CET's dashboard, which automatically processes structured data for credit analysis, fraud detection, loan monitoring, and generates credit memos.

What benefits did CET gain from switching to a single framework?

The benefits include the ability to support multiple models within one framework, more than 50% improvement in efficiency, and a 10% increase in reasoning speed.

What role does the Nvidia GPU play in CET's solution?

The Nvidia GPU supports the processing power needed for CET's intelligent document processing, enabling the handling of larger Language Models and different kinds of GPUs within a single framework.

What is the significance of the Inception program for CET?

The Inception program is invaluable for CET, especially with the APAC team's support, which helps CET grow by introducing different partners and customers, and inviting CET to impactful events.

Can you name two solutions offered by CET as mentioned in the script?

Two solutions offered by CET are 'intelligent document processing' to automate workflows and 'Fast lender workflow automation solutions' for increasing loan books and making fast credit decisions.

How does CET's intelligent document processing solution help customers with unstructured documents?

The intelligent document processing solution helps customers by automating the workflow from intake to processing of unstructured documents.

What is CET's goal in relation to Nvidia?

CET's goal is to become one of the first Language Model (LM) service providers in APAC in close collaboration with Nvidia.