Himanshu Kumar Sharma

About Himanshu Kumar Sharma

I\'m an energy + people style person. I believe in regular improvement, even if it\'s 1% but should be regular.I love writing code because it helps me relate the problem statement in a mathematical fashion.The first code I wrote was for Arduino Uno( microcontroller) where I along with my team were working on ultrasonic sensors and try to use it to make hand gloves which can help specially-abled people detect obstacles and from there the love for coding started.In the initial days I was coding in C, slowly and gradually I picked the Web programming where I along with my team developed a static website with HTML/CSS/JavaSript. Now I code in Python, and since one year I have been working on Machine learning, Deep learning algorithms and the fascinating Natural Language Processing. Even when the office hrs are over I tend to continue writing codes. Most recently I have been working on building DNN which can detect the 2D images as a rectangle or circle wherein I used famous fgg16 application using Keras to preprocess the images.This is what which keeps me light in the night, and ignite the never-ending thrust of creating beautiful, powerful and fascinating things and share them with the world.

Experience

1. Consultant - Ai & Data

   Deloitte

   Sep 2024 — Present · Bengaluru, IN

   PMI (July–Present)Leading Generative AI Initiatives: Led a 5-member team to architect and deliver the Personalized Advanced Learning (PAL) application—implemented advanced GenAI solutions for adaptive, voice-interactive learning experiences.System Architecture & Data Modeling: Designed async LLM response streaming with prompt-caching for rapid query turnaround, including voice-interactive retrieval-augmented generation (RAG) leveraging multimodal speech-to-speech models and podcasting via AWS Polly.DevOps & CI/CD Automation: Initiated and built CI/CD pipelines, authored Dockerfiles for both front-end and back-end, ensuring seamless deployments and scalability.Knowledge Management (KM) Use Case: Developed scalable architecture, data modeling, OpenSearch knowledge base, and upskilled the team on AWS ecosystem with a focus on Bedrock KB, LLM, and Opensearch serverless Vector-DB.Cost Optimization & Best Practices: Partnered on AWS Bedrock cost estimation, implemented GitHub pre-commit hooks, and drove adoption of robust coding standards.Nestlé (Jan–July)RAG Application Delivery: Built and deployed an advanced RAG-based application with high-performance document processing (PDF, PPT, DOCX, XLSX), automated ingestion pipelines for SharePoint, and robust logging frameworks.Containerization & Automation: Spearheaded end-to-end containerization (Docker Compose), Azure shared volume setup, and automated CI/CD deployments.Team Enablement: Drove continuous learning through proactive knowledge sharing and guided POCs in voice-to-voice RAG and graph-based RAG approaches.Leadership & Impact HighlightsLed cross-functional teams on cloud-native, GenAI, and knowledge management initiatives.Mentored peers on AWS Bedrock, OpenSearch, and best DevOps practices.Guided architecture and deployment for cutting-edge RAG, voice-AI, and multi-modal AI solutions.Fostered a culture of continuous improvement and operational excellence.

Education

• BITS Pilani Work Integrated Learning Programmes

  Master of Technology - MTech, Data Science & Engineering

• SHRI RAMSWAROOP MEMORIAL COLLEGE OF ENGINEERING AND MANAGEMENT, LUCKNOW

  Bachelor's degree, Electronics and Communications Engineering

  2014 — 2018

• Jaycees Public School

  intermediate, Science

  2012 — 2013

• Jaycees public school

  10th, Science

  2010 — 2011