CS Honors @ UMass Amherst • AI/ML Researcher (Time-Series focused) • Incoming SDE Intern @ Philips
I build AI systems, data pipelines, and research-driven ML tools that solve real problems with clarity and strong engineering.
- 🔬 AI/ML Researcher building tooling, datasets ingestion & evaluation pipelines for Time-Series Foundation Models
- ❤️ Excited to work on the Patient Monitoring Solution at Philips during my Co-op, where my knowledge and work will directly improve patient outcomes!
- 🏋️♂️ Dedicated to bodybuilding and fitness, focusing on discipline and self-improvement with a goal to compete one day!
- 🇷🇺 Actively learning Russian, and pursuing it as a secondary major at UMass
- ☕ Always open to talking about AI research, health, languages, or building meaningful technology in the era of AI!
Languages:
Python · JavaScript · TypeScript · C# · C · SQL · HTML/CSS
Frameworks & Tools:
PyTorch · TensorFlow · NumPy · Pandas · FastAPI · Node.js · Docker · AWS (Lambda, S3, ECR) · Git · React
ML/Research Tools:
LoRA · HuggingFace · Lightning · Chronos/MOMENT/Papagei/Mantis · Scikit-Learn · Jupyter
Toolkit integrations, dataset loaders, and evaluation utilities for Chronos/MOMENT/Papagei/Mantis across forecasting, classification, and anomaly detection.
Secure ML pipeline using guardrails, AWS Lambda, Docker microservices, and automated document ingestion.
A modular computer-vision training & inference framework with plug-and-play dataset configs and augmentation presets.
AI-driven flood/police/traffic simulation with agentic AI workflows and scenario management.
- Benchmarking TSFM models on more diverse domains and new tasks (imputation soon!)
- Adding LoRA adapters and multi-task training to TSFMs
- Scaling agentic workflows & ML pipelines on my Crisis Simulation Engine project
- Prepping for UMass DataFest in Spring 2026!
LinkedIn: LinkedIn
Email: Oboukantar@umass.edu