Phanindra Parashar
About Candidate
Data Scientist with approximately 4 years of experience in ML, predictive modeling, and data analysis. Proficient in deep learning (PyTorch, TensorFlow), NLP, and time-series forecasting (XGBoost, CatBoost). Skilled in leveraging AWS (Sagemaker, Lambda, API Gateway) and MLOps (Docker, MLFlow) for model optimization and scalability. Collaborates effectively with cross-functional teams, translates complex business requirements into data-driven solutions, and communicates insights to stakeholders. Committed to driving business value through cutting-edge data science techniques and staying at the forefront of industry trends.
Location
Location
Education
• 1.4 GCPA (German Scale) • Advanced Machine Learning (1.0 German Scale) • Machine Learning (1.0 German Scale) • Predictive Modelling (1.7 German Scale)
Work & Experience
• Machine Learning & Revenue Optimization: Developed revenue optimization solution using advanced ML techniques (XGBoost, CatBoost) for time-series forecasting. Built ETL pipeline for feature engineering and ranking algorithms, driving 16% ad revenue growth (Confirmed results with A/B Testing) across 1200+ websites, benefiting 35+ publishers. • Cloud Computing & Scalable Architecture: Deployed ML models using AWS tech stack (Sagemaker, Lambda, API Gateway, EC2), optimizing performance for 25M daily users. Leveraged Docker and asynchronous programming to streamline data preprocessing efficiency in large-scale environment. • Data Analysis & Stakeholder Collaboration: Partnered with cross-functional teams to identify optimization opportunities, translating complex business requirements into mathematical models. Analyzed 85TB data using BigQuery, developing anomaly detection algorithms. Built Locker Studio dashboard to monitor model performance and deliver actionable insights. • NLP & Team Collaboration: Collaborated with data science teams on NLP applications for text classification, achieving 73% accuracy on 26 categories using open-source embeddings. Actively participated in Scrum calls and knowledge transfer sessions to foster teamwork and drive successful outcomes.
• Machine Learning & Deep Learning: Developed hierarchical reinforcement learning algorithm for autonomous driving using PyTorch. Implemented and tested innovative algorithms, showcasing expertise in deep learning frameworks (PyTorch, TensorFlow). Optimized neural networks through hyperparameter tuning. • Research Collaboration & Communication: Collaborated with a large research team, effectively communicating complex ideas and presenting findings. Merged traditional control systems with reinforcement learning and created proof-of-concept simulations to demonstrate framework adaptability and effectiveness.
• Stakeholder Collaboration: Collaborated with local government and palm oil companies to optimize logistics and reduce costs. Built strong relationships with stakeholders to drive project success. • Adaptability & Leadership: Demonstrated adaptability by pivoting business model to better meet market needs. Led and coordinated teams to develop innovative solutions in a challenging environment.
