Portfolio - Yasodha C

2

Machine Learning
Projects

1

Deep Learning
Project

2

Python
Projects

1

GenAI
Project

About

Hello! I'm Yasodha, a Machine Learning Engineer with hands-on experience building production-ready AI systems through internships and self-directed projects.
I work comfortably with Python, SQL, PyTorch, Scikit-learn, and modern deployment tools to deliver end-to-end solutions — from data preprocessing and feature engineering to model training, evaluation, and API integration using FastAPI and Streamlit.
Currently, I focus on Generative AI, developing RAG pipelines with ChromaDB, Groq/Llama models, Text-to-SQL agents, and semantic routing to create intelligent applications like e-commerce chatbots.
I approach problems methodically, prioritize clean and scalable code, and learn quickly to turn ideas into working solutions.
I'm ready to join a collaborative team and contribute to real-world AI products from day one. Let's connect!

Key

Skills

Python

SQL

Statistics

Machine Learning

FastAPI

Deep Learning

Model Deployment

Data Analysis

My

Projects

E-Commerce Chatbot

Domain/Function: E-Commerce

An intelligent customer support chatbot for e-commerce platforms that handles both general policy questions and dynamic product searches using Generative AI.

Car Damage Detection

Domain/Function: Computer Vision

AI-powered Car Damage Detection model that identifies front/rear damage types using a fine-tuned ResNet50 architecture, reaching ~80% accuracy. Includes advanced augmentation, Optuna tuning, and a Streamlit interface for real-time predictions—supporting faster, more reliable insurance evaluations.

Credit Risk Prediction System

Domain/Function: Finance

AI-powered credit risk prediction system using XGBoost and Logistic Regression on 50K+ records to assess loan default probability. Applied SMOTE-Tomek and Optuna tuning, achieving AUC 0.98 & KS > 85. Deployed an interactive Streamlit dashboard with FastAPI for real-time scoring.

Health Insurance Premium Prediction

Domain/Function: Health Care

Developed a health insurance premium prediction model using Linear, Ridge, and XGBoost with extensive EDA and feature engineering. Implemented age-based model segmentation and introduced a genetic risk feature to significantly improve prediction accuracy and model stability.

Atliq Hotels Data Analysis

Domain/Function: Hospitality

Performed a comprehensive data analysis of AtliQ Hotels using Python and multiple datasets to uncover insights on revenue trends, occupancy, customer behavior, and booking platforms. Applied data cleaning, EDA, and visualization techniques to generate actionable business recommendations.

Expense Tracking System

Domain/Function: FinTech

Developed a full-stack Expense Tracking System using FastAPI, Streamlit, and MySQL, enabling users to record, update, and analyze expenses with category-wise visual insights and automated backend logging.

My

Experience

AI/ML Intern – Labmentix Pvt Ltd
May 2025 – Nov 2025 | Remote

Designed, trained, and optimized predictive machine learning models (regression, classification, tree-based ensembles) using Scikit-learn, XGBoost, and LightBGM for business use cases.

Built complete ML pipelines: data ingestion with SQL queries, cleaning, preprocessing, feature engineering, and model evaluation — achieving up to 15% gains in key metrics (AUC, F1, RMSE).

Performed Exploratory Data Analysis (EDA) using Pandas, NumPy, and Seaborn/Matplotlib to discover trends and deliver actionable insights to stakeholders.

Applied cross-validation, hyperparameter tuning (GridSearchCV, Optuna), and model versioning using Git/GitHub for reproducible experiments.

Developed and deployed interactive Streamlit dashboards for real-time predictions, feature importance, and business reporting.

Maintained clean, collaborative codebase on GitHub with proper branching, pull requests, and documentation.

Data & Operations Analyst - Prakash Cottage Works, Coimbatore

Aug 2019 – Feb 2021 | Full-Time

- Analyzed daily production data to track and report output of 50–100 commercial wet grinder stones per day, building summary reports, trend visuals, and performance dashboards in Excel to identify bottlenecks and support process improvements.

- Managed inventory and procurement tracking through structured Excel spreadsheets, forecasting material requirements, monitoring stock levels, and minimizing shortages or overstock to maintain smooth production.

- Handled financial and compliance data management in Excel, including invoicing, expense recording, billing reconciliation, and GST documentation preparation; automated recurring reports and improved data accuracy and turnaround time.

- Maintained centralized vendor and supplier trackers in Excel, analyzed delivery performance and timeliness patterns, and generated summary insights to optimize sourcing decisions and reduce production delays.

Associate Software Engineer – Fidelity Investments, Bangalore
Apr 2015 – Feb 2017 | Full-Time

Monitored and supported 100+ critical batch workflows weekly on mainframe systems, ensuring 99.9% on-time delivery.

Optimized batch performance by ~20% through SQL query tuning, resource profiling, and process re-engineering.

Automated recurring reports using JCL, REXX, and SQL stored procedures, reducing manual effort by ~60%.

Led large-scale data cleanup initiatives using advanced SQL scripts and archiving strategies, improving system efficiency and cutting storage costs.

Collaborated with global teams (India, US, Ireland) on incident management and process standardization following ITIL practices.

Hello, I am

Yasodha C

2

1

2

1

About

Key

My

E-Commerce Chatbot

Car Damage Detection

Credit Risk Prediction System

Health Insurance Premium Prediction

Atliq Hotels Data Analysis

Expense Tracking System

My

Awards

Contact Me

Email

Github

Linkedin

Call

Yasodha C

Download Resume