TATA CAPITAL LIMITED 

 

1.Name of the Customer: TATA Capital Limited (TCL) 

TATA Capital Limited (TCL), is one of India’s leading NBFCs, offering a wide range of  financial products and services for individuals, SMEs and corporate clients. 

2.Challenges Faced by the Customer:  

Currently, TCL users who use the TCL Loan Management System (LMS) spend a lot of time reaching out to the SPOCs of the LMS to debug the issues they face while using the LMS. This often leads to delays in receiving responses from the SPOCs.  

3.Why AWS and Why ACC? 

AWS GenAI provides a powerful and flexible platform for developing and deploying AI and ML applications. With its comprehensive set of tools, scalable infrastructure, seamless integration with other AWS services, and strong security measures, it is a robust choice for organizations looking to harness the power of artificial intelligence. Whether you’re building sophisticated models or integrating AI into existing applications, AWS GenAI can help you achieve your goals efficiently and effectively.  

ACC had several successful GenAI projects from the Banking and Financial Services sector under its belt.   

4.Solution Provided by ACC:  

• ACC built a ChatBot for TCL, to improve user satisfaction for TCL LMS, with a focus on optimizing information retrieval processes and ensuring real-time access to essential data.  

• This chatbot provided users with accurate, contextually relevant, and coherent responses by combining the strengths of information retrieval systems with generative AI capabilities.   

• We have used the Cohere model from AWS Bedrock for Retrieval-Augmented Generation (RAG) and stored vector data in an RDS Postgres database, which supported storing vector embeddings. For the ChatBot (Question Answering) feature, we have utilized the Claude Sonnet 3 LLM model from AWS Bedrock. Amazon Titan Text Embeddings is used to convert text to vector data.

5.AWS Services used:

1. AWS Bedrock 
2. Amazon Elastic Kubernetes Service (EKS) 
3. Amazon Elastic Compute Cloud (EC2) 
4. Amazon Relational Database Service (RDS) for PostgreSQL
5. Amazon Elastic Container Registry
6. Application Load Balancer
7. AWS Lambda
8. Amazon Titan Text Embeddings
9. Amazon Simple Queue Service (SQS)
10. Amazon CloudFront

6.Results and Benefits: 

• The TCL ChatBot provides round-the-clock support, allowing users to get answers and assistance at any time, even outside regular banking hours. 

• It offers immediate responses to user queries, reducing wait times and improving overall user satisfaction 

• By automating routine inquiries and tasks, chatbots reduces the need for a large customer service team, leading to cost savings on staffing and training. 

• Automated responses reduce the risk of human error in handling routine transactions and inquiries. 

• Thus, implementing a chatbot in the TCL environment delivers numerous benefits, including enhanced user service, cost efficiency, and improved operational efficiency. By automating routine tasks and providing 24/7 support, chatbots not only improve customer satisfaction but also enable TCL to optimize its resources and streamline their operations.

7.Metrics for Success: 

• Website’s information access will be more streamlined and user-friendly, providing customers and employees with an enhanced experience. 

• Employees will experience improved efficiency as they gain quick and convenient access to the information they need, reducing manual search efforts. 

• The integration of AI assistance will significantly decrease the chances of errors in information retrieval and dissemination. 

• The implementation of this generative AI solution- ChatBot should transform the way information is accessed and shared, creating a more efficient and dynamic environment for all stakeholders.

8.Lessons Learnt: 

• Thoroughly understand and map out the common questions and issues that users have. A chatbot should address the most frequent and significant user needs to be effective. 

• Regularly gather and analyze user feedback to refine and enhance the chatbot’s performance and functionality. 

• Maintain consistency in the information provided by the chatbot and human agents to avoid confusion and provide a coherent customer experience.

9.Start Date of the Engagement: 21st May 2024  

10.End Date of the Engagement: 31st July 2024 

11.Architecture Diagram: 

About ACC: 

ACC is an AWS Advance Partner with AWS Mobility Competency. Awarded The Best BFSI industry Consulting Partner for the year 2019, ACC has had several successful cloud migration and application development projects to its credit. Our business offerings include Digitalisation, Cloud Services, Product Engineering, Big Data & Analytics and Cloud Security. ACC has developed several products to its credit. These include Ottohm – Enterprise Video and OTT Platform, Atlas API – API Management and Development Platform, Atlas CLM – Cloud Life Cycle Management, Atlas HCM – HR Digital Onboarding and Employee Management, Atlas ITSM – Vendor Onboarding and Service Management and Smart Contracts – Contract Automation and Management.