Shail K Patel

My Research

Research work in machine learning and data-driven solutions

A Two-Stage, Leakage-Aware Framework for Early Academic Risk Detection in Undergraduate Engineering Cohorts

Sep 2025

Proposed and implemented a leakage-aware, two-stage ML framework to forecast student academic decline
Applied leakage-aware modeling to ensure predictions are temporally valid and unbiased
Evaluated 50+ regression pipelines per subject (≈200 total) to benchmark academic performance forecasting methods
Tested 50+ classification pipelines for risk detection; best-performing stacking ensemble achieved recall 0.657
Applied explainability (SHAP) to support model interpretability and trust

Student Performance Prediction Early Warning Systems Ensemble Learning Stacked Classification Leakage Prevention Model Explainability (SHAP)

Read Paper →

My Projects

Innovative solutions powered by Machine Learning

PredictGrad

Jul 2025

Built a full ML pipeline to predict Semester 3 marks and detect academic risk
Regression: Voting Regressor (Ridge + Lasso + ElasticNet)
Classification: Stacking (CatBoost, LGBM, ExtraTrees)
Deployed as interactive Streamlit app with XAI explanations

Python Scikit-learn SHAP Streamlit

Try It Live →

Falsifi.AI

Aug 2025

Verifies scanned signatures as genuine or forged using deep learning-based handwriting analysis.
Runs ~5 experimental pipelines combining distinct preprocessing and model architectures for robustness.
Signer-disjoint train/val/test on the CEDAR dataset to evaluate on unseen identities.
Streamlit UI for uploading an anchor and a candidate; outputs label, confidence, and GradCAM/SHAP explanations.

Python PyTorch CNNs ResNet TorchVision

Try It Live →

RaddiWala

Aug 2025

Digitized the traditional scrap-selling process by connecting customers with Raddiwalas (scrap collectors).
Enabled photo-based pickup requests, competitive bidding, and trust-driven mutual ratings.
Role-based workflows: Customers upload scrap photos and receive bids; Raddiwalas place offers and manage pickups.
Passwordless Email + OTP authentication with JWT sessions and secure cookies.
Integrated Multer for image uploads. Nodemailer for OTP delivery.

Node.js Express.js MongoDB (Mongoose) Pug Vanilla JS JWT Multer Nodemailer

GitHub Repo →

Beyond The Marks

Mar 2025

ML pipeline with explainable AI to analyze student performance
Detects grading bias and assesses influence of sensitive attributes
Quantified teacher effectiveness with performance indicators
Statistical analysis of attendance, grades, and student features

Python SHAP Pandas Matplotlib

Try It Live →

Attend2Achieve

Mar 2025

Analyzed subject attendance and marks correlations
Performance pattern explanation with statistical measures
Data visualization with plots and heatmaps
Academic insights dashboard with improvement suggestions

Python Pandas Seaborn Streamlit

Try It Live →

The Profit Oracle

Feb 2025

Financial analysis and revenue forecasting tool
Automated insights on returns and customer behavior
Location-based analysis with ML algorithms
Cost vs price comparison and profit analysis

Python Scikit-learn Streamlit NumPy

Try It Live →

Work Experience

Professional journey and achievements

GetMySpace

Machine Learning Engineer

Jan 2025 – Jun 2025

GetMySpace is a parking management solutions startup using AI and automation to streamline vehicle and space tracking.

Developed GetMySpace's first AI-powered prototype for real-time parking management, integrating computer vision and machine learning to automate vehicle and slot detection
Built and deployed models using Python and PyTorch, reducing end-to-end processing time to ~4 seconds and achieving real-time update latency of ~0.7 seconds
Designed and implemented backend logic and MongoDB-based systems to handle prediction outputs, user logs, and dynamic updates, ensuring scalable and efficient performance

Technologies Used

Python PyTorch Computer Vision Machine Learning MongoDB Backend Development

Skills & Expertise

Technologies I work with

Education

Academic journey and learning path

Bachelor of Engineering in AI & ML

2023 - 2027

Java Software Engineering Mathematics-1 Physics IoT Computer Hardware

Advanced Java PostgreSQL MySQL Mathematics-2 Data Structures Electronics

Web Development Python Probability & Statistics

Theory of Computation Computer Organization Discrete Mathematics MERN Stack Supervised ML Advanced Python

.NET Aptitude Economics & Finance

Let's Work Together

I'd love to hear about your project and how I can help with ML solutions.

I'm always excited to work on innovative projects that make a real impact. Whether you need ML solutions, data analysis, or want to explore AI possibilities, let's connect and create something amazing together.

Send Email LinkedIn GitHub