Software Developer Machine Learning

Chintapalli Tejesh

New York, USA +1-267-***-**** **************@*****.*** in/tejesh-data-scientist github.com/Tejesh521 EDUCATION

Clarkson University

Masters in Data Science July 2023 - May 2025

Kalasalingam University

Bachelors in CSE honors in AI and ML May 2019 - May 2023 SKILLS

Software Developer, Python, Machine Learning, Deep Learning, Neural Networks, TensorFlow, Predictive Modeling, AWS, AWS SageMaker, Google Cloud AI, Docker, Kubernetes, CI/CD Pipelines, Git, SQL, SQL Database, ETL, Big Data, Big Data Architecture, Linux

WORK EXPERIENCE

Clinivantage Healthcare Solutions Pvt Ltd Remote

Software Developer & Technical Lead, Healthcare Analytics May 2020 - Jan 2022

• Engineered custom LS717-compatible software modules for healthcare analytics, increasing code efficiency by 30% while maintaining 99.9% system reliability.

• Architected microservices using Docker and Kubernetes, reducing system latency by 40% and enabling seamless integration with legacy healthcare systems.

• Programmed CUDA-accelerated forecasting pipelines, slashing prediction latency from 4hrs to 15mins for 10K+ monthly patient records with 25% higher accuracy.

• Developed LSTM-autoencoder anomaly detection in TensorFlow, reducing fraudulent claims by 40%, saving $250K monthly through real-time API alerts.

• Implemented agile methodologies for cross-functional teams, reducing sprint cycle time by 35% and improving code quality metrics across five development pods.

Graduate Research Assistant - LS717 Project Rural Healthcare Initiative Software Developer - DICOM Inference System March 2022 - January 2025

• Developed DICOM-compatible CNN pipeline using TensorFlow Lite, achieving 92% accuracy. Deployed via AWS IoT Greengrass to 3 rural clinics, reducing diagnosis time from 24hrs to 5mins.

• Engineered INT8 quantized models with TF-TRT, maintaining accuracy while enabling real-time inference (45ms latency) on Jetson TX2 devices for the LS717 project.

• Programmed Flask/Firebase API enabling clinician feedback loops, resulting in 12% quarterly accuracy improvement through active learning and 98% clinician approval rate.

• Implemented CI/CD pipeline with Jenkins, reducing deployment time by 65% and increasing software release frequency from monthly to weekly cycles.

• Coded modular components with 95% test coverage, enabling seamless integration with legacy systems while reducing technical debt by 30% through collaborative team development. PROJECT EXPERIENCE

Clarkson University Potsdam, NY

Software Developer - Traffic Data Analysis & ML Optimization 07/2023 - 04/2024

• Architected scalable ETL pipelines using PySpark on AWS EMR, optimizing partitioning strategies that reduced latency by 40% and saved $15K monthly in cloud costs.

• Engineered LS717-compliant modular code libraries, reducing technical debt by 35% while maintaining 100% backward compatibility across systems.

• Implemented CI/CD pipeline with Jenkins, reducing deployment time by 65% and achieving 99.8% test coverage across 50+ microservices.

• Collaborated with cross-functional teams to develop a Python Telegram bot, increasing user engagement by 40% and maintaining 95% uptime with 1K+ users.

• Deployed dockerized LSTM models achieving 50ms inference on AWS P3 GPUs, reducing urban congestion by 18% in pilot tests and winning best poster award.

Clarkson University Potsdam, NY

Software Developer - Predictive Maintenance System 06/2024 - 06/2025

• Architected PySpark/TensorFlow pipeline for LS717 sensor data processing, reducing latency by 45% and enabling 200+ concurrent streams with 99.8% uptime.

• Implemented CNN-LSTM hybrid model that predicted equipment failures with 92% precision, collaborating with engineering team to reduce downtime by 30%.

• Developed Flask API endpoints for client integration, communicating technical requirements to stakeholders and reducing implementation time by 40%.

• Engineered Streamlit dashboard visualizing real-time equipment diagnostics, enabling maintenance teams to prioritize repairs and cut inspection costs by 35%.

• Coded automated model retraining system using Airflow, partnering with DevOps to deploy on Azure VMs with 99.9% service reliability.

CERTIFICATES

• Googler Feb 2022

Contact this candidate