Hi, I am Swati Hegde

I turn complex problems into elegant AI solutions and help to understand why models behave the way they do. I build GenAI systems, automate repetitive workflows, and keep products simple — even when the underlying technology is complex.

About Me

About Me

I am an AI/ML professional with 7+ years of experience building data-driven and GenAI solutions that deliver real-world impact. I specialize in developing and optimizing machine learning and generative AI models, creating insightful visualizations, and uncovering actionable insights. I enjoy tackling challenging projects that drive meaningful outcomes and am passionate about designing scalable, reliable AI and predictive modeling solutions. Continuously learning and exploring new technologies, I strive to stay at the forefront of the evolving data and generative AI landscape.

⚙️ Technical Skills

Programming Languages: Python, SQL and PL/SQL

Machine Learning & Deep Learning: PyTorch, TensorFlow, Keras, Scikit-learn, NLP, CNNs, RNNs

Generative AI and Agentic AI: LLMs, LLaMA, Gemini 2.5 Flash, Agentic AI, Prompt Engineering

ML Techniques: Classification, Regression, Clustering, Recommender Systems

Application Development & MLOps: FastAPI, Flask, Streamlit, Git, Model Deployment

Cloud Technologies: AWS - S3, EC2, SageMaker, Dynamo DB, RDS

Data Processing & Analytics: NumPy, Pandas, Matplotlib, Seaborn, Tableau

My Projects

AccessJobs – AI-Powered Career Assistant

This project implements AccessJobs, an intelligent job search and career preparation platform for tech professionals and emerging scholars. It fetches recent job postings based on user criteria and uses LLMs via the Groq API to provide personalized fit scores, semantic gap analysis, and tailored interview preparation.

LLM(Llama 3.3 70B) Prompt Engineering python-jobspy

CineMatch – Personalized Movie Recommendation System

This project implements a Hybrid Movie Recommendation System using the MovieLens dataset. The system combines Collaborative Filtering (user–item rating patterns) and Content-Based Filtering (movie metadata such as genres and tags) to deliver more accurate and engaging recommendations.

Recommender systems Collaborative Filtering

FraudShield – Credit Card Fraud Detection System

Fraudulent credit card transactions are a major challenge for the financial industry, leading to significant losses and reduced customer trust. FraudShield is a machine learning project that leverages Logistic Regression, Random Forest, and XGBoost to detect fraudulent transactions with high accuracy.

Logistic Regression Random Forest XGBoost

RetailVision – U.S. Retail Sales Forecasting

RetailVision is a time series forecasting project that predicts U.S. monthly retail sales using Prophet, SARIMA and Linear Regression to uncover seasonal trends, anomalies, and growth patterns.

Time Series Linear Regression Prophet SARIMA

Get in Touch

I’m always open to new opportunities, collaborations, or project discussions. Feel free to reach out and drop me a message!