Data Scientist with Python, SQL, and Cloud Deployment

Programming: Python, SQL

DBMS: PostgreSQL, NeonDB

Data Wrangling and ML: Pandas, NumPy, TensorFlow, Scikit- learn

ETL & Deployment: Docker, GitHub, GCP

Data Visualization & Analytics: Matplotlib, Seaborn, Tableau, PowerBI

Data Scientist with hands-on experience in data engineering and machine learning projects. Skilled in Python, SQL, and PostgreSQL, with exposure to deep learning models [CNN, LSTM]. Experienced in deploying containerized apps with Docker and GCP, SUMMARY

+91-859******* · *************@*****.*** · @mylinkedin · @mygithub Hyderabad, Telangana, India

Hemanth Jayan

SKILLS AND TECHNICAL PROFICIENCY EDUCATION

Integrated MSc Data Science Sep 2020 - May 2025

Amrita Vishwa Vidyapeetham, Coimbatore CGPA: 7.1 CBSE Higher Secondary May 2017 - Apr 2019

Mary Mount Public School, Kottayam Percentage: 95% CBSE Primary and Secondary Mar 2006 - Apr 2017

Our Own English High School, UAE CGPA: 10.0

PROFESSIONAL EXPERIENCE

Vishal Peripherals Secunderabad, Hyderabad Jun 2025 - Present Data Science Intern

Architected a PostgreSQL pricing & rebate engine with versioned price lists, effective-dated joins, and aging SKU detection; delivered per-line rebate & Net Landing Cost views with stacking and ad-hoc support tracking across brands—reducing rebate- amount discrepancies in brand claims by 38% through validation and guardrails. Built and shipped a full stack [PERN] Lead Management System for retail and corporate sales: a React (Tailwind) app with role-based dashboards for managers and sales reps, unified lead detail/assignment, activity history & reminders, quoting/proposals, and pipeline analytics. Implemented secure REST APIs (Node/Express, JWT) on NeonDB with audit trails and strict ownership/RBAC; migrated corporate proposal storage to Google Cloud Storage. Deployed via Docker/Nginx on Cloud Run—resulting in durable file links, no cross-manager data exposure, and smoother lead operations. National Remote Sensing Centre Balanagar, Hyderabad Aug 2024 - Jan 2025 Research Intern

Leveraged knowledge of deep learning and geospatial techniques to create a model capable of accurately predicting Land Use/Land Cover from satellite data.

Implemented U-Net architecture with a ResNet backbone that returned a pixel classification prediction rate of 93%. Utilized QGIS, a geographic system software to work with raster data and vector polygons to create raw data to feed into model. PROJECTS

Gender Classification Using CNN Deep Learning

Developed a convolutional neural network (CNN) model to classify gender from image data. Utilized a dataset of images with preprocessing techniques for data augmentation and splitting. Built a CNN architecture with layers including Conv2D, MaxPooling2D, Dense, and Dropout for improved generalization. Achieved 86% accuracy, validated on test data using TensorFlow and Keras. Forecasting US Domestic Automobile Inventory Levels: A Comparative Study Deep Learning Engineered and compared 10 forecasting models including ARIMA/SARIMA, LSTM, GRU, CNN-LSTM hybrid, and ensemble methods using Python.

Built a novel CNN-LSTM hybrid architecture that achieved 75% variance explanation and 27.05 RMSE, dramatically outperforming traditional ARIMA models.

Integrated 12+ macroeconomic indicators including unemployment rates, consumer sentiment, exchange rates, and recession indicators to enhance prediction accuracy.

Implemented advanced statistical validation using ADF and KPSS tests, autocorrelation analysis, and proper temporal train-test splits.

Natural Language Processing with Sequence Models Coursera, March 2024 - May 2024 COURSES AND CERTIFICATIONS

Contact this candidate