Ashish Kumar

I build predictive models and data pipelines using Python, SQL, and cloud analytics tools to turn complex datasets into actionable insights.

Through Google Cloud Data Analyst training, I've built ELT pipelines in BigQuery, created dimensional data models, and designed Looker dashboards and LookML semantic models for analytical decision-making.

I've developed time-series forecasting models for Tesla stock prices, benchmarking ARIMA against RNN and LSTM architectures and achieving an R score of 0.963. I also built a movie recommendation system using NLTK, cosine similarity, and KNN, plus a CNN-based CAPTCHA recognition pipeline that reached 78.04% accuracy.

At IIT Kanpur, I applied systematic analysis to improve a vertical-axis wind turbine's operational efficiency from 18% to 33%. My aerospace engineering background, distributed-computing projects, and experience mentoring over 100 Dynamics students strengthen my quantitative and problem-solving approach.