Turning Shipping Data into Route-Level Insights
Analyzed 10,000+ shipping records to uncover route-level performance patterns across factories, states and ship modes.
- Python
- Pandas
- Plotly
- Streamlit
Data Science × Analytics × AI
3rd-year BCA Data Science student with a minor in Data Analytics, currently working as a Data Analyst Intern and exploring neural networks.
Currently working
Data Analyst Intern @ Unified Mentor
Exploring
Neural Networks
I work where data, technology and curiosity meet.
I'm a 3rd-year BCA Data Science student with a minor in Data Analytics. Alongside academics, I'm gaining practical experience as a Data Analyst Intern at Unified Mentor while exploring neural networks through research, experimentation and implementation.
Two internships, one direction of study. Open a role to see what the work involved.
A map of the areas I've worked in and am moving toward — not a straight line from one role to the next.
Current role. Work areas include:
Going deeper into how models learn, alongside my degree.
Featured internship project
I recently worked on a project with Unified Mentor analyzing factory-to-customer shipping performance for Nassau Candy Distributor.
An internship analysis project — not a role in the company's logistics operations.
Abstract representation. Not actual route or geographic data.
Without route-level visibility, logistics decisions stay reactive.
Self-directed study and experiments — not published research.
Exploring the foundations and behavior of neural networks through study, experimentation and implementation.
Hover, focus or tap a skill to see what it connects to.
Select a skill
Connections light up between skills that I use together.
One internship case study and one prototype I built myself.
Analyzed 10,000+ shipping records to uncover route-level performance patterns across factories, states and ship modes.
An AI-assisted maternal health risk assessment web application prototype designed to support risk identification and healthcare workflows.
A prototype for maternal health risk assessment.
Maternal health risk is judged from a small set of routine readings. This prototype asks a practical question: can a trained model turn those readings into a clear Low, Mid or High risk level?
A Flask web app wraps a trained Random Forest classifier. A user enters six health readings and gets a predicted risk level along with a probability for each category.
Python and Flask serve the app. A Random Forest model with standardized features does the prediction. A JSON endpoint, /api/predict, allows programmatic access. The interface is plain HTML, CSS and JavaScript.
Illustrative sketch of the form flow, not a screenshot.
A learning prototype for exploring how ML can support risk identification in healthcare workflows. It is for educational and research purposes only. It has not been clinically validated and is not a diagnostic tool.
View the repositoryBCA — Data Science
MIT-ADT University
Minor Data Analytics
Interested in data, AI, meaningful problems and building things that actually work.
Or write to suryawanshianuj83@gmail.com