Chitresh Kaushik
Summary
AI and Data Science professional with five years of experience, including three years in Generative AI, NLP, and Time Series Forecasting. Skilled in building end-to-end AI solutions, with expertise in Agentic RAG, LLM-based reasoning, and intelligent workflow automation. Proficient in predictive modeling, data-driven insights, and scalable deployment of AI systems to solve complex business problems.
Overview
8
8
years of professional experience
Work History
Data Scientist
Medtronic (IBM)
11.2024 - Current
- Designed and deployed Agentic RAG-based chatbot system for multilingual customer care, improving case resolution efficiency by 45%.
- Developed GenAI-powered automated email generation reducing manual responses by 70% across US and EU markets.
- Integrated Azure Speech SDK for real-time voice interaction, enabling 25% faster response cycle.
- Built and maintained scalable Azure Functions with 99.9% uptime.
- Improved multilingual categorization F1-score from 0.76 — 0.89 via prompt optimization and data augmentation.
Data Scientist
IBM Consulting
12.2022 - 10.2024
- Owned end-to-end Time Series Forecasting Pipeline with 10+ models deployed, achieving 12% improvement in accuracy.
- Built Reinforcement Learning framework that reduced forecasting error (MAPE) by 15%.
- Delivered executive-level analytics enabling data-driven talent planning, impacting $10M+ in workforce optimization.
- Designed ethical AI model evaluation pipeline for responsible AI monitoring.
- Drove integration of Bayesian inference methods, reducing forecast volatility by 18%.
Data Scientist Associate
Wipro Technologies
02.2018 - 07.2019
- Built and deployed predictive maintenance models using Python (Scikit-learn, Pandas, NumPy) to identify early equipment failure, reducing downtime by 22%.
- Developed customer churn prediction model using logistic regression and random forests, improving retention strategy accuracy by 18%.
- Designed data preprocessing pipelines for 1M+ transactional records, ensuring high data quality and 99% feature consistency across training sets.
- Automated model evaluation and hyperparameter tuning workflows using GridSearchCV, reducing manual effort by 60%.
- Deployed trained models via RESTful APIs using FastAPI, integrating seamlessly with client-side dashboards.
Education
M.Tech - CS
IIT Dhanbad
Dhanbad, Jharkhand07-2022
B.Tech - CS
Amity University
Panchgaon, Haryana07-2017
Skills
- Retrieval-augmented generation systems
- LLM optimization techniques
- LangChain and LangGraph frameworks
- Agentic AI development
- Model deployment and APIs
- Prompt evaluation strategies
Timeline
Data Scientist
Medtronic (IBM)
11.2024 - Current
Data Scientist
IBM Consulting
12.2022 - 10.2024
Data Scientist Associate
Wipro Technologies
02.2018 - 07.2019
M.Tech - CS
IIT Dhanbad
B.Tech - CS
Amity University