Computer vision research
Pothole Severity Detection
A YOLOv9 instance-segmentation system for detecting road damage and assigning severity levels from images and video.
YOLOv9PyTorchFastAPI
Available for ML and AI roles
Raj Bariya.
Machine Learning Engineer
I build machine learning tools across computer vision, transportation, and applied language models—from data pipelines to usable interfaces.
Baltimore, MarylandResearch & engineering
Research / Development
Selected work
Applied LLM system
Automated Sports Reporting
A structured generation pipeline that converts match statistics and player data into journalist-style reports and player summaries.
PythonOpenAINLP
End-to-end machine learning
AI Flight Delay Prediction
A prediction pipeline combining historical flight and weather data with a Flask API and an interactive interface.
TensorFlowFlaskSQLite
Frontend engineering
Portfolio Website
A responsive portfolio focused on accessible navigation, clear project evidence, and fast delivery across devices.
Next.jsTypeScriptTailwind CSS
View repository for Portfolio WebsiteWhere I’ve worked
Experience
Aug 2025 — Present
Undergraduate Research Assistant
CEAMLS DEPA Lab, Morgan State University
- Built a YOLOv9 instance-segmentation pipeline that classifies potholes by severity in images and video.
- Deployed the computer-vision workflow through FastAPI with color-coded annotations for faster road inspection.
- Developed an LLM pipeline that turns structured match and player data into publication-ready sports reports.
YOLOv9FastAPIPythonPyTorchLLMs
May 2025 — Aug 2025
Research Intern
CEAMLS SAIRI, Baltimore, MD
- Developed a prediction system over more than eight million records.
- Built internal web applications that made model outputs usable by non-technical stakeholders.
- Reworked data-processing bottlenecks and measured 50x to 200x faster execution across the optimized steps.
XGBoostFlaskAirflowETL
A little background
About me
I enjoy the full path from an idea to a dependable tool: cleaning data, evaluating models, designing APIs, and making the result useful.
My recent work focuses on road-infrastructure computer vision and automated sports reporting. I care about measurable performance, clear communication, and systems that survive beyond a demo.
The toolkit
Skills & tools
- Python
- C++
- Java
- NumPy
- Pandas
- scikit-learn
- PyTorch
- TensorFlow
- XGBoost
- LLMs
- SQL
- Docker
- AWS
- FastAPI
- Airflow
Get in touch
Let’s talk.
I’m open to machine learning, AI engineering, and research opportunities. Tell me what you are working on and where I can help.
rajbariya1256@gmail.comBaltimore, Maryland