NSaiHarshithVarma
Data Scientist @ Confluent
Author of ML from Scratch
Data Scientist at Confluent, building AI-driven systems for product analytics. Author of ML from Scratch, a deep-dive series on deriving machine learning algorithms from first principles.

Expert
Kaggle Rank
Top 3%
23
Notebooks
500+ Forks
2
Research
Published Papers
5+
Projects
Open Source
Vol. 1
Book
373 pages · ML from Scratch
01Selected work
02Current focus
MALLORN: Detecting Tidal Disruption Events
Developing a machine learning solution for the MALLORN Challenge to classify Tidal Disruption Events (TDEs)—stars torn apart by black holes. The system analyzes simulated LSST photometric lightcurves, utilizing advanced time-series feature extraction to identify rare transients within highly imbalanced astronomical datasets.
- Data Science
- Feature Engineering
- Imbalanced Classification
- Time Series Analysis
