Data Scientist & ML Engineer with 6+ Years of Experience

SUMMARY

Data Scientist and ML Engineer with *+ years of experience designing end-to-end ETL pipelines, building predictive ML models, and developing computer vision systems using Python, PyTorch, CUDA, SQL, and OpenCV. Strong expertise in data cleaning, feature engineering, time-series analysis, deep learning, GPU- optimized inference, and scientific computing. Experience working with large-scale meteorological data, neuromorphic event-based data, and geospatial analytics. Published researcher and hands-on problem solver delivering production-ready solutions across government, research labs, and enterprise analytics environments.

Seeking roles in: Data Science • Machine Learning • AI Research • Data Engineering • Computer Vision CHAITANYA CHAUDHARY

Harrison, NJ +1-862-***-**** ************@*****.*** Linkedin DATA SCIENTIST • MACHINE LEARNING ENGINEER • ETL ENGINEER (6 YEARS EXPERIENCE) TECHNICAL SKILLS

Programming: Python, SQL, C++, Java, Bash

ML & AI: Scikit-Learn, PyTorch, TensorFlow, CUDA, MoveEnet, XGBoost, Random Forest ETL & Big Data: Pandas, NumPy, HDF5, MapReduce, batch/stream processing Computer Vision: OpenCV, stereo geometry, triangulation, pose estimation Cloud & DevOps: Docker, Git, Linux, REST APIs, Flask Analytics & Viz: Matplotlib, Seaborn, D3.js, Tableau (basic) Concepts: Predictive Modeling, Time-Series Forecasting, Optimization, Research Methodology, Statistical Analysis

Machine Learning Engineer Jan 2024 – Present

R Systems International — Applied AI & Robotics Lab California Processed neuromorphic DVS event streams and built a real-time data preprocessing pipeline for pose estimation tasks.

Designed multi-view batch ETL workflows (aggregation, format transformation, cleaning, alignment) using NumPy + HDF5.

Performed GPU-accelerated inference using MoveEnet, improving throughput by 32% via CUDA optimizations.

Built a 3D reconstruction pipeline using stereo geometry + triangulation to map elbow/wrist/shoulder joint coordinates into world coordinate space. Delivered 3D motion trajectories to the Robotics Lab for downstream real-time motion prediction and force-assisted control modules.

Documented reproducible experiment pipelines enabling researchers to scale model training across datasets.

Tech: Python, PyTorch, CUDA, NumPy, OpenCV, HDF5, DHP19 PROFESSIONAL EXPERIENCE

Data Scientist / Machine Learning Engineer Aug 2020 – Dec 2023 Indian Meteorological Department (Govt. of India) New Delhi, India Designed, built, and optimized large-scale ETL pipelines to ingest, validate, clean, and transform multi-terabyte weather radar datasets from 30+ stations across India. Improved short-term forecasting accuracy by 18% by implementing noise reduction, interpolation, and temporal smoothing algorithms.

Automated generation of radar-based visualizations such as MAX-Z, reflectivity, velocity, and turbulence used by regional forecasting centers.

Applied time-series forecasting models for rainfall & storm prediction, incorporating spatial/temporal correlations.

Developed anomaly detection scripts that identified extreme weather signatures with 92% detection precision.

Collaborated with scientists to integrate processed datasets into legacy systems and internal research platforms.

Presented model findings and visualization pipelines to meteorology leadership teams. Tech: Python, Pandas, NumPy, Matplotlib, ETL, Time-Series, Interpolation, Clustering Data Analyst / Junior ML Engineer Jan 2019 – Jul 2020 Skylar Softech Ahmedabad, GJ

Developed predictive models for various businesses including sales forecasting, user churn prediction, and pricing analytics.

Analyzed client datasets, conducted preprocessing and complex data cleaning (missing value imputation, outlier removal, normalization).

Implemented classification and regression models achieving 10–25% improvement over baseline accuracy.

Built reusable ETL pipelines, automated reporting scripts, and interactive visual dashboards

(Matplotlib, Seaborn, D3.js).

Performed A/B testing and statistical evaluations for marketing conversion analytics. Tech: Python, SQL, Scikit-Learn, Matplotlib, Pandas Applied Research — Event-Based Hand Pose & Trajectory Modeling Rutgers University, 2024 – Present

Processed DVS event streams into frame representations for upper-limb pose estimation. Implemented cleaning, multi-view merging, temporal batching, and file structuring in HDF5. Built GPU-accelerated inference pipeline with PyTorch + MoveEnet. Created 3D joint reconstruction using stereo geometry & triangulation. Supported robotics team with real-world trajectory outputs for downstream prediction models. RESEARCH & PUBLICATIONS

Publication

Model-Free Gait Recognition using Random Forest

Procedia Computer Science (Elsevier), ICMLDE 2023

Proposed entropy-based feature extraction & model-free gait classification. Rutgers University — New Brunswick, NJ

Master of Science in Computer Science

Coursework: Data Visualization, Statistics, Advanced Algorithms, Data Management EDUCATION

Pandit Deendayal Energy University — India

B.Tech in Computer Engineering (2019 – 2023)

Coursework: DS & Algorithms, OS, Cloud Computing, Networks, Software Engineering

Contact this candidate