The customer chose to protect their confidential information
AI Multi-Channel Assistant for Banking
About the client
The customer was a prominent bank operating in Kazakhstan with a strategic focus on mortgage products.
![[object Object] on the map](https://content.andersenlab.com/common-data/client-maps/kazakhstan/1741687566.png)
About the project
The customer engaged Andersen to build a scalable AI-powered multi-channel Assistant to optimize client interactions across all applicable channels. The system required multi-channel intent recognition, integration with RAG-driven knowledge bases, and the use of AWS AI services for text recognition, document parsing, and intelligent search across internal and external datasets.
Challenges
The bank needed a modern AI solution capable of handling high query volumes, complex document retrieval, and multilingual support while maintaining security and cost efficiency. Key challenges included:
- LLM optimization on AWS, including fine-tuning, prompt engineering, and domain-adapted datasets to deliver accurate chatbot interactions.
- Multi-channel delivery, ensuring seamless user experiences across digital platforms through scalable AWS back-end services.
- Cost-optimized AI operations, achieved by selecting the most efficient AWS AI/ML models and tuning resource usage to match projected traffic and query patterns.
Solution
Andersen engineered and deployed an AI-powered multi-channel assistant on AWS, seamlessly integrated into the bank’s digital channels. The resulting solution leverages multiple LLMs, including Claude, Cohere’s models, and HuggingFace’s embeddings, while providing REST APIs for text query processing and RAG-based knowledge management. The team designed an optimized storage architecture to support multi-channel delivery and enhanced the chatbot’s behavior through structured prompt engineering for more accurate content retrieval. The system is deployed on a robust AWS foundation and utilizes the following AWS services, among others:
- Amazon Bedrock to engineer and scale generative AI apps with foundation models;
- Amazon API Gateway to manage secure API requests, providing for seamless interactions between the chatbot and external solutions;
- AWS Lambda for serverless execution of chatbot logic, handling real-time text processing and query routing;
- AWS CloudFormation to automate infrastructure provisioning, a critical requirement for scalable and efficient chatbot deployment;
- Amazon OpenSearch Service to enable fast and relevant search queries, enhancing the chatbot’s ability to retrieve information from large datasets.
App functionality
Delivered AWS-based functionalities include:
- Automated client support, enabling users to receive immediate AI-driven responses across digital channels;
- Personalized assistance powered by LLMs and contextual data for tailored guidance;
- Intelligent document search using RAG pipelines and AWS AI services for high-accuracy retrieval;
- Smart loan advisory, providing users with relevant mortgage and financial insights based on predefined business rules and AI reasoning;
- Seamless integration with human agents, allowing escalation to bank specialists when needed.
Project results
Outcomes include:
- Substantial query throughput: roughly 2 million queries were handled in the first month;
- Better response accuracy: chatbot interactions have been improved across multiple digital channels;
- Robust security and compliance: client data is protected via AWS KMS, AWS Secrets Manager, and full PCI DSS compliance;
- Accelerated deployment: rollout times have been reduced from days to hours thanks to CI/CD pipelines and AWS CloudFormation;
- Increased user engagement, owing to the flawless multilingual support (English, Kazakh, and Russian are available);
- Monthly infrastructure costs have remained stable due to AWS auto-scaling and pay-as-you-go pricing;
- Pay-as-you-go and auto-scaling options have lowered expenses by about 40% compared to on-premises solutions.
About Andersen
The project was delivered by Andersen’s AWS-certified engineers, who specialize in building cloud-native solutions and implementing secure and scalable architectures aligned with AWS Well-Architected Framework principles.
Let's talk about your IT needs
What happens next?
An expert contacts you after having analyzed your requirements;
If needed, we sign an NDA to ensure the highest privacy level;
We submit a comprehensive project proposal with estimates, timelines, CVs, etc.
Customers who trust us
