Generative AI Assistants Enhance Productivity and Personalize the Self-Service Employee Experience

By Manish Madhukar, Client Partner – Cognizant
By ManojKumar Kotakonda, Technical Lead – Cognizant
By Bala KP and Deepak Singh, Sr. Partner Solutions Architects – AWS
By Jasmine Rasheed Syed, Sr. Customer Solutions Manager – AWS

Cognizant

One of the key challenges organizations face is managing and disseminating knowledge in an efficient manner. Traditionally, accessing relevant information or guidance within an organization is a time-consuming process as employees may need to search through extensive documentation, engage in lengthy email exchanges, or rely on internal experts for answers to their queries.

These highly manual processes can hinder decision-making and negatively impact productivity. Furthermore, as remote work and globalized teams become more normal, the need for instant access to accurate and tailored information has become more pressing than ever.

The rise of artificial intelligence (AI) assistants can help companies address this knowledge management challenge. Digital assistants powered by generative AI are capable of promptly delivering relevant and personalized responses to cater to the distinct knowledge requirements of every individual in a company. However, the true potential of AI assistants extends beyond simple information retrieval.

AI assistants have the capacity to reshape workplace efficiency, streamline processes, reduce response times, and empower employees with the tools they require to accelerate productivity. This sets the stage for organizations to explore and harness the transformative power of generative AI, not just as chatbots but as intelligent assistants capable of enhancing every aspect of work.

This post discusses how Cognizant, an AWS Premier Tier Services Partner and Managed Service Provider (MSP), built an AI assistant solution for a human resources (HR) support function to enable a personalized self-service experience for employees. The solution uses Amazon Bedrock, a fully managed service to build and scale generative AI applications with foundational models, and Amazon Kendra, an enterprise search service that’s powered by machine learning (ML) on Amazon Web Services (AWS).

Solution Overview

Amazon Bedrock offers a choice of high-performing foundation models (FMs) from leading AI companies like AI21 Labs, Anthropic, Cohere, Meta, Stability AI, and Amazon via a single API. It provides a broad set of capabilities to build generative AI applications, simplifying development while maintaining privacy and security.

Cognizant’s solution in this post uses Anthropic’s Claude v2 large language model (LLM) for providing the conversational question and answer (Q&A) experience. A web-based chat application is built and hosted in AWS Amplify which sends user queries to Amazon Bedrock through Amazon API Gateway and an AWS Lambda function. Amazon Cognito is used for authentication and access control for the application, while Amazon Kendra is used for indexing company’s internal and external web content.

To limit the AI assistant to provide responses using only company-specific data, Cognizant uses a technique called Retrieval Augmented Generation (RAG). With this approach, the solution retrieves information that are most relevant to users request from the enterprise content indexed in Amazon Kendra.

The solution then bundles that information as context along with users’ requests as a prompt sent to the Claude v2 FM running on Amazon Bedrock. LangChain, a publicly available framework for building generative AI applications, is used for orchestration between Amazon Kendra and Amazon Bedrock.

Figure 1 – AI assistant reference architecture.

The workflow includes the following steps:

1. 1. Create an Amazon Kendra index and use the web crawler connector v2.0 to crawl and index company website and associated contents.
   2. User accesses AI assistant application hosted in AWS Amplify.
   3. User is redirected to Amazon Cognito-hosted user interface (UI) login page for authentication and is provided with a Q&A chat interface upon successful authentication.
   4. Chat application issues an API call to Amazon API Gateway based on the user’s query.
   5. API Gateway invokes LangChain orchestration code in AWS Lambda to send the query as a search request to Amazon Kendra.
   6. Lambda sends the search query to the Amazon Kendra index.
   7. Kendra returns the results to the Lambda orchestrator.
   8. The Lambda LangChain orchestrator includes user query along with the retrieved information from Kendra as context in the LLM prompt and sends to Amazon Bedrock.
   9. Bedrock returns the user’s response to the Lambda function.
   10. Response is returned to user through Amazon API Gateway.
   11. User gets the requested response for the query.

Use Case Walkthrough

Let’s take a human resources AI assistant use case to show how this tailored AI assistant can work. The data considered for demonstration purposes is publicly available careers information from the company website. The HR AI assistant addresses the need for an efficient and user-friendly solution to provide comprehensive and up-to-date information about careers in an organization.

Currently, information about an organization’s work culture, job openings, and various other aspects of employment is scattered across multiple sources, making it challenging for users to access the information they need quickly and accurately. When a user asks a question to the HR AI Assistant, it searches through information and returns the relevant information from the user’s query and provides a succinct response back to the user.

Figure 2 – User interacting with AI assistant.

Solution Benefits

Organizations can leverage AI assistants to create personalized knowledge hubs:

• Enhanced workplace efficiency: AI assistants can reduce the time employees spend searching for answers from internal repositories. This efficiency improvement can lead to faster decision-making and improved productivity across the organization.
• Tailored knowledge: AI assistants are designed to cater to specific knowledge needs. This leads to a better quality of work as employees receive information that’s more relevant to their roles and tasks.
• Reduced workload on experts: By automating responses to common queries, AI assistants alleviate the burden on subject matter experts within the organization. Experts can focus on more complex and strategic tasks, leading to a more efficient use of their skills.
• Improved employee experience: Access to AI assistants creates a more satisfying work environment by empowering employees with readily-available resources, which can lead to increased job satisfaction and higher morale.
• Versatility across industries: The adaptability of AI assistants mean they can be applied across various industries and roles. Whether you’re in IT, HR, management, or any other line of business, there’s a solution that can fit your unique needs.
• Cost-efficiency: AI assistants can reduce the cost associated with employee training and support, as they can handle routine inquiries and leave human experts to focus on more specialized tasks.
• Competitive advantage: Organizations that adopt AI assistants can gain a competitive edge by streamlining operations, improving responsiveness, and staying at the forefront of technological innovation.

Cognizant built an AI assistant solution for transforming an HR support function. The AI assistant enabled self-service experience offering rapid responses to HR queries, automating routine tasks, and offering personalized career guidance.

Conclusion

This post shows the transformational impact of AI assistants equipped with tailored organization-specific knowledge. By proactively addressing the challenges of managing diverse knowledge needs within a growing organization, enterprises can effectively harness the power of AI-driven solutions.

Cognizant continues to assist clients in unlocking these potential avenues, and its forward-thinking approach exemplifies the transformative potential of AI-driven solutions in reshaping how organizations manage and deliver specialized knowledge.

.

.

Cognizant – AWS Partner Spotlight

Cognizant is an AWS Premier Tier Services Partner and MSP that transforms customers’ business, operating, and technology models for the digital era by helping organizations envision, build, and run more innovative and efficient businesses.

Contact Cognizant | Partner Overview | AWS Marketplace | Case Studies