Mahantesh Pattadkal

Bengaluru

Summary

Generative AI Engineer with extensive experience in developing and deploying production-ready AI applications. Specializing in Generative AI, NLP-based products, and Retrieval-Augmented Generation (RAG) systems to enhance automation and streamline decision-making. Expert in advanced data science, Explainable AI (XAI) techniques, and creating executive dashboards to track KPIs and demonstrate the direct impact of AI initiatives on business growth.

Overview

years of professional experience

Work History

Gen AI Engineer

Baxter International

03.2024 - Current

Led the development of Agentic AI applications tailored for multiple departments within Baxter, enhancing automation and decision-making processes.
Built and deployed enterprise-wide Gen AI products, allowing company-wide audiences to leverage AI for day-to-day tasks, thus boosting productivity.
Designed and implemented Generative AI products for the finance team, significantly elevating their efficiency and accuracy in financial reporting.
Developed end-to-end NLP-based products for customer support teams, enabling faster query categorization and resolution, thereby improving customer satisfaction.
Implemented Retrieval Augmented Generation (RAG) systems for the products team to assist with product queries, providing quick and accurate information retrieval.

Data Scientist

KNIME GmbH

05.2020 - 02.2024

Developed and deployed advanced data science models using KNIME and Python on AWS and KNIME Server, utilizing MLOps for efficient lifecycle management.
Utilized Explainable AI (XAI) techniques to provide clear insights into model behavior and facilitate stakeholder understanding.
Created executive dashboards to track essential KPIs, aiding business decision-making and driving growth.
Conducted data science trainings for the customers and their data science teams.

Education

Master of Science - Digital Engineering

Otto-von-Guericke University

07.2022

Bachelor of Technology - Electronics & Communication

Vellore Institute of Technology

07.2015

Skills

Programming Languages: Python, R (Basic), MySQL, DAX
NLP Frameworks: NLTK, spaCy, Hugging Face Transformers, BERT, GPT

Gen AI Frameworks: LangChain, LangGraph, LlamaIndex, TensorFlow, Keras, PyTorch
Database: SQL Server, PostgreSQL, Qdrant, Redis, Pinecone

Projects

Baxter Agents Studio
• Developed an enterprise-level platform enabling departments to create customizable agentic RAG systems, allowing users to drag
and drop documents for generating accurate responses..
• Implemented extensive customization options, including choice of Large Language Models (LLMs), customizable prompts, and multilanguage
support.
• Built the backend using LangGraph, FastAPI, and PostgreSQL, with Qdrant database for efficient vector storage.
• Developed the frontend with Streamlit, providing a user-friendly interface for seamless interaction.
• Implemented Retrieval Augmented Generation (RAG) systems Designed for enhanced productivity and ease of use in an enterprise
environment.

Agentic RAG for Finance CXOs Baxter International
• Developed an Agentic RAG system using LangGraph and LlamaIndex, enabling finance CXOs to obtain real-time answers to their
queries.
• Enhanced response accuracy with metadata filters and a feedback loop, and included citations to build user confidence
• Utilized Redis for efficient storage and retrieval of embeddings, ensuring quick and reliable access to information.
• Conducted prompt tuning and leveraged LightRAG for graph-based responses, catering specifically to the finance department’s needs.
• Deployed a live chatbot for a year, assisting CXOs with daily queries and promoting self-reliance in decision-making.

Timeline