R
Ravinandana Appoji
About
Detail
Gujarat, India
Résumé
Jobs
verified_user
0% verified
AI/ML EngineerTempusOct 2024 - Current (2 years 1 month)Engineered CNN models with TensorFlow and PostgreSQL, applying Regression and Predictive Analysis, reducing radiologist errors by 37%, increasing throughput by over 1,800 studies per month, saving approximately $450K annually. Implemented NLP pipelines leveraging GPT-4 and Claude with RAG to extract structured clinical insights, cutting manual review by around 15 hours per week, increasing report consistency by 22%, and accelerating urgent diagnosis turnaround. Optimized YOLOv8 and Detectron2 pipelines with Kafka streaming, improving critical case detection by 28%, reducing review latency by 10 minutes per study, and enhancing triage accuracy for 1,500+ urgent workflows monthly. Applied SHAP and LIME for interpretability, managed experiment- AI/ML EngineerDeloitteJan 2022 - Dec 2023 (2 years)Architected XGBoost and CatBoost ML frameworks to refine KPI forecasting, increasing prediction accuracy by 22%, reducing client portfolio risk by $2.5M, and improving enterprise operational decision-making efficiency. Automated document classification using NER, Azure Databricks, and MySQL, reducing misclassification errors by 28%, accelerating compliance workflows for nearly 1.5M pages per month, and streamlining regulatory reporting. Integrated CI/CD pipelines via GitHub Actions, Docker, and Kubernetes on Azure ML, cutting deployment failures by 12%, reducing release cycles by 35%, and saving Deloitte approximately $250K annually in cloud infrastructure costs. Monitored ML pipeline health using Datadog and Azure Monitor, ensuring 99.5% u
- Associate Machine Learning EngineerShopifyJan 2021 - Dec 2021 (1 year)Developed recommendation systems using PyTorch and Neural Networks, improving personalization accuracy by 24%, increasing conversion rates by 3%, and generating $1.1M additional revenue across 3.2M+ user sessions. Built preprocessing workflows with Pandas, NumPy, and MongoDB, reducing feature extraction errors by 18%, optimizing real-time training pipelines, and efficiently handling more than 400k daily transaction records. Applied Text Classification and Summarization with BERT, detecting misclassified sentiment, improving product ranking relevance by 21%, and enhancing monthly product recommendation accuracy across the catalog. Created evaluation dashboards in Jupyter Notebook, visualizing hyperparameter impacts, improving retraining effi
- ML InternLexisNexis Risk SolutionsSep 2020 - Dec 2020 (4 months)Built fraud detection models using scikit-learn, reducing false positives and successfully identifying over 2,000 potentially fraudulent transactions across multiple datasets. Processed structured datasets using PostgreSQL in VS Code, ensuring 99% data integrity, minimizing errors, and supporting ML pipelines for 200k historical transactions. Developed Speech Recognition and streaming analytics via Apache Kafka, automating voice verification, reducing misidentification by 24%, and processing 5k+ calls weekly.
Education
verified_user
0% verified
- Master in Information TechnologyUniversity of North TexasJan 2024 - May 2025 (1 year 5 months)Denton, TX
- Bachelors of Engineering in Electrical and Electronics EngineeringRV College of EngineeringJun 2018 - Jun 2023 (5 years 1 month)India
- AAWS Certified Machine Learning– Specialty
- OOracle AI Vector Search Professional
- AAWS Certified Data Analytics- Specialty
- MMachine Learning Specialization-Coursera