DEVSHREE JADEJA
+1 934-263-1865 · devshreehjadeja@gmail.com · linkedin · ORCID · Portfolio · GitHub
EDUCATION
Stony Brook University, M.S. in Computer Science (Data Science) GPA: 4.0/4.0 Aug 2024 – Expected May 2026
Coursework: Statistics, Qualitative Research , Computer Vision, Natural Language Processing
Pandit Deendayal Energy University, B.Tech in Computer Science GPA: 9.91/10 Jul 2020 – May 2024
Coursework: Data Structures and Algorithms, Operating Systems
WORK EXPERIENCE
ISRO(Indian Space Research Organisation) Jan 2024 - Jun 2024
Research Intern
• Constructed CNN/RNN models to decode satellite imagery, boosting classification accuracy by 18% across 5+ terrain
types. Spearheaded processing of 12M+ image tiles, improving throughput by 22% using Apache Spark and SQL.
• Implemented deep learning pipelines in TensorFlow/Keras, achieving 93% accuracy in cloud and anomaly detection.
• Overhauled map reporting with automation tools, cutting delivery time by 47% via Python and Tableau. Secured
real-time ingestion of 400K+ satellite images using Kafka-based ETL pipelines, improving accuracy by 25%.
• Collaborated on deploying 3 ML pipelines on AWS with Docker, maintaining 99.5% uptime for scalable analytics.
• Partnered with scientists to integrate XAI techniques and statistical models, enhancing classification transparency by
21% across 4 satellite data types, while communicating findings through cross-functional discussions.
DigiWagon Technologies May 2023 - Aug 2023
Data Science Intern
• Built a web portal using PHP and Django, streamlining data access & retrieval speed by 30% using BFS algorithms.
Implemented Redux for asynchronous API calls, boosting engagement by 50% and reducing load times by 25%.
• Engineered data extraction and Promoted to Lead Intern within 4 months for exceptional performance in data analysis.
• Incorporated SQL to manage and query 500K+ records, supporting data integrity and backend optimization.
• Executed A/B testing for feature rollout, accelerating product decisions and reducing time-to-market by 25%.
Shanro Key Chem Jan 2021 – Nov 2021
Machine Learning Intern
• Crafted CI/CD pipelines using Scala and Postman reducing deployment errors by 30% & improving release speed.
• Architected auto-scaling backend for 2.5M+ API requests, cutting server costs by 15% and reducing latency by
200ms. Trained XGBoost and Decision Tree models for property prediction, achieving 89% accuracy.
• Optimized features and hyper parameters using Scikit-learn and Optuna, reducing training time by 20%.
PROJECTS
FinSense: Real-Time Financial Recommendation Engine | LLM, GenAI Mar 2024 - Jun 2024
• Orchestrated a financial engine with Apache Spark, MapReduce, AWS S3, and DynamoDB, processing 5M+ data
points/second, reducing latency by 35%, and enabling data visualization through Tableau dashboards.
• Initiated a BERT-based LLM using PyTorch, enhancing recommendation relevance by 30% and satisfaction by 25%.
StreamLineAI – Intelligent ML Pipeline for Ad Click Prediction | MLOps, PySpark, AWS Oct 2023 - Dec 2023
• Ingested and scheduled 10M+ records using Kafka, AWS S3, GCP Storage, and Airflow, boosting pipeline efficiency
by 35%. Validated and transformed data with PySpark, improving data quality metrics by 28%.
• Applied XGBoost, LightGBM, and CatBoost models, increasing CTR prediction AUC by 22%. Deployed real-time
API with FastAPI, Docker, and CI/CD on AWS EC2, reducing latency by 40%.
AgriDetect: Potato Disease Detection System | FastAPI, CUDA, CNN Jan 2022 - Mar 2022
• Trained a CNN in TensorFlow for image corruption detection with LIME and Grad-CAM visualization.
• Compressed and converted the model to TensorFlow Lite, reducing size by 72% and inference time by 45%. Boosted
training speed by 3.5× using CUDA GPUs and mixed precision, lowering memory usage by 38%.
SKILLS
Programming & Scripting: Python, R, SQL, Java, Scala, Java Script, MATLAB, Swift, GO, Julia
Data Analysis & Visualization: Tableau, Power BI, Matplotlib, Seaborn, Excel, D3.Js, Plotly, Looker
Machine Learning & AI: TensorFlow, PyTorch, Scikit-Learn, Pandas, SKLearn, Deep Learning, CoreFlow ,
Transformers, GAN, LLM, RAG, Natural Language Processing (NLP), Computer Vision, GenAI, CodeLLM
Big Data Technologies: Apache Spark, Hadoop, Kafka, NoSQL Databases, Hive, Pig, BigQuery
Deployment & Cloud Technologies: AWS, GCP, Azure, Docker, Kubernetes, Airflow, Jira
