Summary
Overview
Work History
Education
Skills
Personal Information
Projects Executed
Timeline
Generic

SHRUTI HONRAO

Pune

Summary

AI Algorithm Engineer with proven expertise in generative AI and real-time data streaming at Barclays Global Service Centre. Developed solutions that enhanced customer service efficiency and agent productivity. Proficient in AWS technologies, optimizing workflows to deliver measurable results through strategic collaboration. Experienced in Agile and Scrum methodologies, consistently achieving project goals with scalable code and effective team management.

Overview

6
6
years of professional experience

Work History

AI Algorithm Engineer

Barclays Global Service Centre
Pune
12.2022 - Current
  • RAG-Based Knowledge Kit for Customer Service Agents: Implemented a Retrieval-Augmented Generation (RAG)-based Knowledge Kit that is dynamically generating contextual responses for customer service agents. This system is reducing information retrieval time by 40%, while enhancing response accuracy, agent productivity, and service efficiency.
  • Real-Time Knowledge Streaming with WebSocket and ECS: Built a real-time streaming solution using WebSocket protocol and Amazon ECS to ensure low-latency knowledge retrieval for customer service agents. This architecture is delivering instant, continuous access to relevant information, thereby improving agent responsiveness and customer experience.
  • Automated Contact Center Solutions with Amazon Connect and Lex: Designed and developed intelligent contact center workflows using Amazon Connect and Amazon Lex, integrating custom Lambda-based business logic, Net Promoter Score (NPS) tracking, and seamless context switching between intents. This solution is automating customer interactions, while enhancing personalization and scaling service delivery.

Software Engineer

Accenture
Pune
02.2020 - 12.2021
  • Conversational AI Development: Spearheaded the development of an intelligent chatbot interface designed to curate topic-based navigation and seamlessly route users to relevant FAQ resources, enhancing self-service efficiency and user engagement.
  • Cloud Contact Center Optimization: Led the implementation of an asynchronous chat proof-of-concept to validate Amazon Connect's scalability, successfully demonstrating its capability to manage high-concurrency chat sessions-exceeding five simultaneous interactions per agent.

Education

B.E - Instrumentation and Control

Cummins College of Engineering
Pune
05-2019

Skills

  • Generative AI, natural language processing, retrieval-augmented generation (RAG)
  • Large language model (LLM) integration, AWS SageMaker, Hugging Face (all-mp-net-base-v2), LangChain, vector data store
  • Amazon Lex, Amazon Connect
  • ECS (Elastic Container Service), API Gateway, Lambda Functions
  • Python
  • AI Agent
  • Amazon Bedrock
  • Lambda
  • CloudWatch
  • context switching between intents
  • Git
  • Retrieval-augmented generation
  • Real-time data streaming
  • AI chatbot development
  • Contact center workflows
  • Cloud computing
  • Software optimization

Personal Information

Title: Generative AI Engineer

Projects Executed

  • Agentic AI for IBAN Retrieval Automation (PoC), Developed a proof of concept using Amazon Bedrock Agent APIs to trigger external API calls at the start of a chat session., Built a Streamlit UI to interact with users and retrieve account and IBAN data., Enabled faster and more accurate responses without using Lambda or Lex.
  • Intent Classification Enhancement for Amazon Lex (PoC), Built a PoC to improve fallback handling in Amazon Lex., Integrated the all-mpnet-base-v2 model from Hugging Face using SageMaker and ECS., Achieved more accurate classification of low-confidence utterances, reducing fallback triggers and improving conversational accuracy.
  • Fallback Intent Handling Using LLM (Haiku Model) (PoC), Built a fallback PoC using the Haiku model via Amazon Bedrock to classify user intents in ambiguous or low-confidence scenarios., Enhanced Lex's ability to handle unclear inputs and improved overall conversational accuracy.
  • Medical Chatbot using GenAI, Developed a medical chatbot leveraging Retrieval-Augmented Generation (RAG) to provide accurate and efficient responses to medical queries., The system enhanced user interaction by retrieving relevant information quickly and offering precise answers., GitHub - Medical Chatbot
  • MCQ Generator using RAG and OpenAI, Developed an MCQ generator leveraging RAG and OpenAI to automatically generate multiple-choice questions., Streamlined the creation of relevant and accurate questions for various topics., GitHub - MCQ Generator

Timeline

AI Algorithm Engineer

Barclays Global Service Centre
12.2022 - Current

Software Engineer

Accenture
02.2020 - 12.2021

B.E - Instrumentation and Control

Cummins College of Engineering
SHRUTI HONRAO