Transforming Employee Experience: Unveiling Airtel Digital’s GPT-Powered Chatbot for HR & Supply Chain Queries

Rajdeep Singh Chauhan
Airtel Digital
Published in
4 min readJan 8, 2024

In the dynamic realm of digital transformation, Airtel Digital has leapt forward with a homegrown AI chat model powered by GPT technology. This innovative chatbot is designed to answer employee queries on HR and Supply Chain policies, redefining how information is accessed within the organisation through the remarkable conversational capabilities of GPT.

The Origin of Airtel’s GPT-Powered Chatbot

Airtel has been at the forefront of Chatbot technology for the last many years, however with the advancement in GPT, LLM space there was a need to enhance the existing Chatbot for solving problems in a transformed manner. Behind this revolutionary initiative is an intricate dance of Airtel Digital’s expert engineering team, data scientists and GPT technology. The chatbot, based on the GPT architecture, harnesses the prowess of Natural Language Processing (NLP) and Machine Learning (ML), providing employees with a sophisticated and human-like conversational interface for navigating complex policy landscapes.

Conversational Wizardry of GPT

The true enchantment of Airtel’s chatbot lies in its conversational prowess, a testament to the capabilities of GPT technology. No longer constrained by rigid commands, employees can engage in dynamic conversations with the chatbot, enjoying a fluid and natural interaction that mirrors the nuance of human communication.

Natural Language Understanding (NLU):

GPT’s advanced NLU capabilities enable the chatbot to comprehend the intricacies of human language. Employees can pose queries and the chatbot interprets and responds with a depth of understanding that is characteristic of interacting with a knowledgeable colleague.

Contextual Brilliance:

GPT’s contextual awareness is a standout feature, allowing the chatbot to maintain coherence throughout a conversation. Whether an employee delves into HR policies or explores the complexities of Supply Chain protocols, the chatbot adapts seamlessly, ensuring a natural flow of conversation.

Personalisation Beyond Expectations:

Going beyond traditional responses, Airtel’s chatbot, powered by LLM (Large Language Model), learns and adapts to individual preferences and give a touch of uniqueness. This personalisation enhances the user experience, creating a bespoke interaction tailored to each employee.

Multi-Turn Conversations Elevated:

Airtel Digital’s chatbot, engages in multi-turn conversations, transforming the interaction from transactional to conversational. Employees can delve deeper into topics, ask follow-up questions, and receive comprehensive information, fostering a truly interactive and dynamic experience.

Benefits for Employees and the Organisation

The deployment of this GPT-powered chatbot brings a wealth of benefits for both employees and Airtel Digital.

Efficiency and Accessibility Amplified:

Real-time access to information eliminates the need for extensive manual searches, making the chatbot an always-available, instant resource that enhances overall efficiency.

Empowering Employees with GPT:

GPT’s prowess empowers employees with a user-friendly interface for policy inquiries, fostering independence and confidence. This empowerment contributes to a positive workplace culture.

Data-Driven Insights, LLM-Style:

The LLM-powered chatbot can collects valuable data on employee queries, providing Airtel Digital with future roadmap of insights into common concerns and areas for improvement. This data-driven approach supports ongoing optimisation and refinement, setting the stage for continued innovation.

High Level Architecture

Pic Courtesy: AI CEO & Investor Club forum on LinkedIn

Knowledge Base Creation and Document Processing: We began by compiling a diverse range of documents to form a comprehensive knowledge base. These documents then segmented into manageable pieces, readying them for deeper processing and integration.

Embedding Models: Each document segment was then converted into numerical embeddings. These embeddings, representing the semantic meaning of the text, are key to enabling computational understanding and processing.

Vector Database and Indexing: We stored these embeddings in a vector database, where they were indexed for efficient retrieval. This indexing is crucial for the quick and relevant response of the chatbot to user queries.

Chat Interface Integration: The user interaction takes place through the Hive App, where user queries are received and processed. These queries are essential for triggering the chatbot’s response mechanism, contextualising the interaction.

Intelligent Retrieval and Response Formation: The system matches user queries against the vector database to find relevant information. This step ensures that the responses are informed and contextually relevant to the user’s inquiry.

LLMs in Response Crafting: The Large Language Model (LLM) utilises the query and retrieved information to formulate coherent and context-appropriate responses. This showcases the LLM’s capacity to integrate contextual understanding with user queries.

Finally, the response generated by the LLM is presented to the user through the chat interface.

Conclusion:

As Airtel Digital integrates GPT technology into its workplace, the GPT-powered chatbot stands as a beacon of the organisation’s commitment to elevating employee experiences. By combining Airtel’s innovative approach with the conversational brilliance of GPT, the chatbot sets a new standard for accessibility and efficiency in accessing HR and Supply Chain information. Looking forward, the fusion of GPT and employee-centric design promises to redefine workplace interactions, creating a more seamless and delightful experience for all.

Acknowledgements:

Heartfelt gratitude to the brilliant minds who brought this vision to life. The tireless efforts and commitment of the team Rahul, Akash Paul, Rajat Chaudhary, Rajdeep Chauhan under the astute guidance of our esteemed leaders Pradipt Kapoor, Rohit Jain & Sudha Gupta have been instrumental in making this groundbreaking initiative a reality.

This article is co-authored by Rajdeep Chauhan & Rahul from Airtel Digital.

--

--

Rajdeep Singh Chauhan
Airtel Digital

Leading HR Engineering at Airtel Digital, integrating Cloud, iOS, Android, Java, ReactJS, and ERP to drive innovation, efficiency, and performance.