🧠 I build data-driven projects, interactive tools, and AI driven research projects with Python, R, and JavaScript.
Languages: Python, R, Java, SQL, JavaScript
Tools: Git, Git, Docker, Kubernet
Cloud: AWS, GCP, SageMaker, RDS, DynamoDB, Redshift, Spark
Frameworks: PyTorch, scikit-learn, Hugging Face, Gradio, Shiny, Optuna, XGBoost
A Very Accurate Representation of My Day
📚 Learning and integrating **MCP** into real applications
😸 Stuck in playing *While True(): Learn*
No exit condition. Only curiosity.
I finished an MS in Computer Science at Northeastern University (GPA: 4.0), where I spent a lot of time making machine learning models behave and explaining why they sometimes don’t.
I’ve worked in:
- AWS (Amazon) — building generative-AI agents for internal automation, wiring cloud services together, and making systems fail more gracefully
- Healthcare ML (Arcadia.io) — disease-risk forecasting with time-series models, feature engineering, and measurable performance gains (AUC 📈 17%)
- Research (Arcadia.io & Northeastern University) — deep learning for heart-failure prediction using RNNs and Transformers, with top results at Poster Day
- Research (HAI Lab, Northeastern University) — HCI-oriented work on personality inference using LLMs, including data curation, evaluation, and research-facing UI design
Some things I’ve done that survived peer review and demos:
- shipped ML systems to production cloud environments
- tuned models until the metrics stopped yelling at me
- built scallable end-to-end pipelines (data → model → evaluation → UI)
- contributed to open-source time-series ML tooling