Data Scientist Machine Learning

Bhaskar Praveen Palacherla

Data Scientist

******************@*****.*** 571-***-**** VA GitHub

SUMMARY

Data Scientist with 3+ years of expertise in developing and deploying machine learning models, leading data analysis projects, and delivering actionable business insights. Proficient in Python, R, SQL, and C++, with hands-on experience in frameworks like TensorFlow, Keras, PyTorch, Scikit-learn, XGBoost, and LightGBM. Skilled in statistical analysis, predictive modeling, and optimization techniques such as regression, clustering, and neural networks to address complex business problems. Adept at directing large-scale data processing, implementing cloud technologies like AWS and Snowflake, and presenting data-driven insights to stakeholders using tools like Tableau and Power BI. Strong focus on mentoring teams, optimizing workflows, and continuously applying new machine learning advancements to exceed business goals. EDUCATION

Master of Science in Computer Science Aug 2022 – May 2024 George Mason University, Fairfax, VA

Bachelor of Technology in Computer Science May 2017 – May 2021 Bennett University, Greater Noida, India

SKILLS

Methodologies: SDLC, Agile, Waterfall

Programming Languages: C, C++, R, Python, SQL, Java Machine Learning Frameworks and Libraries: TensorFlow, Keras, PyTorch, Scikit-learn, XGBoost, LightGBM Data Processing and Analysis: NumPy, Pandas, SciPy, Matplotlib, Seaborn, ggplot Deep Learning: Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Long Short-Term Memory Networks (LSTM), Generative Adversarial Networks (GAN) Natural Language Processing (NLP): NLTK, SpaCy, BERT, GPT-3, Transformer models Big Data Technologies: Apache Hadoop, Apache Spark, Apache Kafka Databases: MySQL, PostgreSQL, MongoDB

Cloud Platforms: Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform (GCP) Development Tools: Git, Docker, Kubernetes, Apache Airflow Model Deployment: Flask, Fast API, TensorFlow Serving, MLflow Data Visualization: Tableau, Power BI, Plotly, D3.js Mathematics and Statistics: Data Structures and Algorithms, Linear Algebra, Calculus, Probability and Statistics Other Skills: Feature Engineering, Hyperparameter Tuning, Model Selection and Evaluation, Cross-Validation, A/B Testing, Data Cleaning and Preprocessing, Data Mining, Predictive Modeling, Statistical Modeling, Quantitative Analytics, Data Analysis and Modeling, Model Evaluation and Optimization EXPERIENCE

Data Scientist ServiceNow, VA Nov 2023 – Current

• Developed XGBoost and Random Forest models with 85% AUC (25% above baseline) to predict "likely-to- engage" users, improving customer service workflows and increasing engagement by 30%, leading to faster query resolution.

• Designed A/B tests to evaluate the impact of machine learning-driven prioritization in IT service workflows, resulting in a 20% reduction in incident resolution times and improved resource allocation across teams.

• Analyzed service request patterns and applied K-Means clustering to segment users by behavior, enhancing automated workflow routing and improving service desk productivity by 25%.

• Processed and engineered datasets from 20M records weekly using Snowflake and Python, creating scalable datasets for machine learning models that optimized service delivery and reduced workflow bottlenecks by 30%.

• Deployed machine learning models (Logistic Regression, Gradient Boosting) in Databricks and PySpark to prioritize high-impact incidents, reducing ticket resolution time by 25% and improving customer satisfaction scores by 15%.

• Automated retraining of prioritization models using CI/CD pipelines, improving model performance over time and reducing manual intervention in workflow optimization by 40%, leading to continuous service improvements. Computer Vision Engineer Deepsight AI Labs, Noida, India May 2021 – Sep 2022

• Optimized object detection models by quantizing FP32 models to INT8 using Intel's toolkit, reducing inference time and boosting FPS by 35%, enhancing response times in security applications like helmet detection in traffic and mask detection in banks.

• Improved object detection accuracy by refining CNN models through extensive training and testing, resulting in more reliable detection for critical security use cases.

• Containerized the entire platform using Docker, streamlining deployment and making it easier to deliver software to clients, enhancing portability and reducing setup times.

• Deployed the platform on AWS EKS using Kubernetes, ensuring zero downtime during updates with rolling deployments and achieving 40% scalability improvement by leveraging horizontal auto-scaling based on traffic load to camera systems.

• Enhanced neural network performance by 30% through memory optimizations, leading to faster predictions and more efficient resource utilization in real-time computer vision tasks.

• Migrated portions of the platform to Red Hat to meet partnership requirements, ensuring seamless integration and compatibility with new infrastructure.

Data Scientist KPIT, India Jan 2020 – Apr 2021

• Assisted in developing machine learning models using TensorFlow, Keras, and PyTorch to support object detection and sensor data analysis for autonomous driving and vehicle diagnostics, contributing to a 15% improvement in model accuracy.

• Performed data cleaning, manipulation, and analysis using NumPy, Pandas, and SciPy, reducing data preparation time by 20% for projects involving vehicle telematics and predictive maintenance, ensuring timely data availability.

• Supported NLP tasks by implementing models such as NLTK, SpaCy, and BERT for analyzing in-vehicle communications and driver behavior, improving text-based analysis accuracy by 15%, enhancing features in connected vehicle systems.

• Conducted hyperparameter tuning and feature engineering to optimize model performance, boosting results by 20% in projects like electric powertrain performance analysis and ADAS (Advanced Driver Assistance Systems) development.

• Deployed machine learning models using Flask, FastAPI, and TensorFlow Serving, reducing deployment time by 35% for AI-driven solutions in mobility systems and vehicle diagnostics. PROJECTS

GDP Prediction of India: Python, Machine Learning, sci-kit-learn, TensorFlow, ARIMA.

• Led the creation of a supervised predictive model for India's GDP growth, achieving 89% accuracy. Integrated linear regression, MLP Regression, and ARIMA, enhancing model reliability.

• Applied advanced machine learning frameworks such as TensorFlow and sci-kit-learn for model refinement. Implemented data analysis and feature engineering to improve predictive precision, ensuring the model’s utility in economic forecasting and decision-making.

Face Emotion Recognition Mobile App: TensorFlow, Python, Convolutional Neural Networks, Android Studio.

• Designed a Face Emotion Recognition app with 92% accuracy using TensorFlow API, Android Studio, and a Convolutional Neural Network (CNN), mastering neural network algorithms, and with a user-friendly interface. Object Detection Mobile app for blind people: Python, Deep Learning, CNN, TensorFlow Lite, COCO model.

• Created a highly accurate object detection app (96%) for visually impaired users with TensorFlow Lite and the COCO model, enabling real-time recognition.

• Engineered deep learning models for mobile deployment, optimizing size and speed by 20% for real-time detection. Supervised user testing with visually impaired individuals, integrating audio feedback and haptic responses to enhance usability.

Contact this candidate