About me
I am Aaradhya Malaviya, a Computer Science student and Software Developer based in Kanpur, Uttar Pradesh. I specialize in building full-stack applications and AI-driven systems, with a strong focus on Machine Learning, Deep Learning, and Generative AI.
My experience spans across Python, JavaScript, React, and FastAPI, where I have shipped end-to-end applications from architecture to deployment. I have worked on complex projects ranging from DQN-based cognitive radio systems to Computer Vision-based license plate recognition, always focusing on clean, maintainable code and real-world impact.
I am comfortable working with both SQL and NoSQL databases, version control using Git & GitHub, and modern tools like Docker. I enjoy solving complex problems, as evidenced by my 275+ solved problems on LeetCode and 5-star rating on HackerRank.
I am actively seeking opportunities to contribute to development teams building innovative products in the software engineering and AI space.
What I'm doing
-
Software Development
Building robust and scalable full-stack applications using modern frameworks like FastAPI and React.
-
Artificial Intelligence
Implementing Machine Learning and Deep Learning models for Computer Vision, NLP, and Reinforcement Learning.
-
Generative AI
Exploring and building with LLMs, Prompt Engineering, and AI agents using tools like Anthropic and AWS.
-
Data Engineering
Handling complex datasets, performing feature engineering, and optimizing data pipelines for ML models.
Experience & Collaboration
-
Data Science Analyst – Amdox
Engineered NeuralRetail, an end-to-end AI-Powered Sales Intelligence & Predictive Analytics Platform. Processed 15M+ daily transactions, delivering demand forecasts, customer intelligence, churn predictions, and inventory optimization via REST API and interactive dashboards.
-
Data & Feature Engineering
Designed a scalable Lakehouse with Apache Spark & Delta Lake. Integrated Great Expectations for DQ ≥ 98%. Built a Feast Feature Store for point-in-time correct retrieval. Engineered RFM scores, lag features, and external regressors using Polars for high-performance computing.
-
ML & Advanced Modelling
Developed a Prophet + PyTorch Lightning LSTM ensemble (MAPE ≤ 10%) and stacked XGBoost + LightGBM churn classifier (AUC ≥ 0.90). Applied DoWhy causal inference for price elasticity. Implemented global and local interpretability using SHAP TreeExplainer.
-
MLOps & Deployment
Configured MLflow for tracking and Evidently AI for drift detection. Established CI/CD via GitHub Actions. Containerized with Docker and deployed via Kubernetes. Built a high-throughput FastAPI REST API and a 5-page Streamlit dashboard.
