🎩 Cloud Sorcerer | AI Alchemist | DevOps Magician
🎓 Master’s student in Computer Science (AI/ML Specialization) at University at Buffalo
🌍 Open-source contributor & tech blogger – Breaking down complex tech so even my grandma gets it!
🌌 Mastering the art of scaling, wrangling containers, and summoning machine learning models
🔍 On a never-ending quest to automate the boring stuff and deploy the cool stuff
💡 Code. Deploy. Observe. Repeat.
- Developed and optimized software solutions for computer vision and OCR-based applications, enhancing image processing efficiency.
- Built and deployed machine learning models for automated text extraction & classification, leveraging deep learning frameworks.
- Designed REST APIs & scalable data pipelines to integrate image recognition capabilities, improving system performance and computational efficiency.
- Developed & optimized IBM Cloud Code Engine, a serverless platform for deploying containerized workloads, enhancing Kubernetes automation & scaling.
- Designed & maintained scalable microservices in Go, integrating managed cloud services for high-performance backend solutions.
- Automated deployments across environments (Prod, Staging, Dev) using Concourse CI, Tekton & Terraform, ensuring seamless infrastructure as code (IaC) implementation.
- Enhanced system observability with ELK Stack, Prometheus & Grafana**, improving real-time monitoring & debugging.
- Led architecture discussions & troubleshooting, ensuring reliable cloud deployments & reducing downtime.
- Wrote technical blogs on IBM Blogs & Medium on cloud deployment strategies.
- Designed & deployed an AI-powered Travel Quote Manager on IBM Cloud, integrating Watson Assistant for chatbot interactions and Sterling OMS for automated order management.
- Developed machine learning models for personalized travel recommendations, packing lists & itinerary generation, leveraging Cloud Functions for seamless automation.
- Created cloud-native assets for Watson Studio & Watson Assistant, enhancing real-time interaction between cloud services.
- Built & deployed a high-performance cloud-based ETL solution on Azure HDInsight, streamlining data ingestion, processing, & storage.
- Automated data masking & encryption workflows using custom Hive UDFs & Terraform, ensuring secure data handling.
- Optimized big data pipelines, integrating Kubernetes for containerized workload management, enabling efficient hybrid cloud operations.
- Contributed to Shipwright, a Kubernetes-based container build framework simplifying YAML-based image creation.
- Developed an automated cleanup tool for Build & BuildRun retention, reducing node consumption by 25%.
- Project Repo: Shipwright on GitHub
- Built a loan prediction system, achieving 94% accuracy & 0.93 F1-score with advanced feature engineering.
- Developed a web-based application for real-time loan approvals & credit history insights.
- Deployed on cloud infrastructure, ensuring scalability, security, and reliability.
- Project Repo: CreditPredict
- Developed a QA system using transformer models, achieving 85% EM & 91% F1-score for context-aware responses.
- Built an end-to-end pipeline for data preprocessing, model training, and inference, deployed as a scalable API.
- Integrated into a user-friendly interface, enhancing conversational AI and educational applications.
- Designed an ML-based predictive system to track & forecast student performance, incorporating temporal correlations.
- Built a bilayered prediction architecture with ensemble models, improving accuracy & early identification of at-risk students.
- Implemented a modular pipeline for data processing, training, and visualization, enabling data-driven academic insights.
- Project Repo: PerformancePrediction
- Developed an ML pipeline to predict breast cancer occurrence, achieving 100% accuracy with Decision Tree and 94.72% with Random Forest.
- Leveraged cloud services for scalable data storage, model training, and deployment, ensuring efficiency and reliability.
- Automated data preprocessing, training, and evaluation workflows, supporting early detection and assisting healthcare professionals.
- Project Repo: Breast Cancer Prediction
💡 "Leave it better than you found it." – A principle that applies to both codebases and life. Refactoring, optimizing, and improving software is just as important as writing it.
💡 Always open to collaborations, discussions, and exciting new projects! 🚀