Ariz Ahmad
Software Engineer
Software Engineer based in Atlanta, GA, United States. I have strong experience in Mobile development as well as AI/ML, including production ML and GenAI systems. I build scalable data pipelines, LLM-powered applications (RAG, agents), and robust Android solutions with a focus on latency, reliability, and maintainability across the entire lifecycle from development to deployment.
Get in TouchWork Experience
Microsoft
Software Engineer
Feb 2022 — Present- Led onboarding of 15+ enterprise teams data to a large-scale analytics platform supporting downstream ML and AI workloads, migrating 200+ TB of data while implementing enterprise-grade security controls (RBAC, encryption, Azure Key Vault).
- Led adoption of AI-assisted development workflows across a 30+ engineer organization by evaluating LLM tools, defining usage guidelines, and training developers, resulting in a 25–35% reduction in development time and improved code consistency.
- Built and maintained CI/CD pipelines for data and ML-adjacent services using Azure DevOps, enabling automated testing, deployment, and monitoring; reduced release cycles by 60% with 99.9% success across 50+ production releases.
- Designed and optimized Spark-based data pipelines used for analytical and ML use cases, reducing query latency by 40% and enabling faster feature generation for downstream models.
Samsung Research Institute
Software Developer - Android
Jul 2016 — Aug 2019- Developed and optimized Android applications for Samsung Galaxy devices used by 50+ million users worldwide.
- Built features using Kotlin and Java following MVVM and Clean Architecture principles with Jetpack components.
- Improved application stability, maintaining a 99.5% crash-free rate across production releases.
- Reduced app startup latency and memory footprint through profiling and performance tuning.
- Partnered closely with product managers, designers, backend engineers, and QA teams to deliver high-quality user experiences.
- Contributed to full release lifecycle including testing, deployment, monitoring, and post-release performance analysis.
University of Florida
Research Volunteer
Aug 2021 — Feb 2022- Developed deep learning models for protein structure prediction using PyTorch, CNNs, and RNNs on large-scale PDB and UniProt datasets; built automated preprocessing pipelines for 500K+ protein sequences, reducing manual data preparation time by 80%.
- Improved fold classification accuracy by 20% over sequence-based baselines using ensemble methods and hyperparameter tuning, while reducing training time by 40% and maintaining 95%+ cross-validation accuracy for secondary structure prediction.
Featured Projects
Robinhood Clone
Stock watchlist app with a real-time, debounced search and filter feature. Built with Jetpack Compose, Clean Architecture, and MVVM. Fetches live stock data from a public API and displays it in a responsive UI.
Smart Text Summarizer
On-device text summarization app leveraging TensorFlow Lite to instantly condense lengthy articles into clear, actionable summaries. Features an intuitive interface for seamless input and real-time results, ensuring privacy and efficiency without relying on the cloud.
Stock Prediction Platform
Supervised learning-based models to predict the stock price of Microsoft using correlated assets and its own historical data.
Weather Intelligence Service
LLM-powered weather intelligence with advanced RAG (Retrieval-Augmented Generation), real-time insights, and comprehensive evaluation metrics.
About Me
"I believe the best software solutions are built with intentionality and craftsmanship. As an engineer with experience in both AI/ML and Android development, I am committed to more than just technical excellence—I care deeply about system reliability, performance, and code quality. Whether designing robust data pipelines, developing RAG and LLM-powered systems, or building Android apps used by millions, I bring a rigorous, end-to-end approach to every project, ensuring solutions are maintainable, scalable, and production-ready."
