Aabaan Koul

Software and Machine Learning Developer End-to-end systems for classification, vision and live workflows

Computer science student and software developer working across machine-learning and full-stack projects, who previously trained classification models that achieved 98% predictive accuracy.

Highlighted stories

Classification models on a 50,000-record dataset

Modeling and feature pipeline optimization

Aabaan engineered classification models and optimized feature pipelines for a 50,000-record dataset. The models achieved 98% predictive accuracy.

Optimizes models and feature pipelines

National Institute of Technology Srinagar

Machine Learning Intern |

Handwritten arithmetic recognition

A vision pipeline for handwritten equations

Aabaan developed a pipeline to segment and evaluate handwritten equations. The work achieved 94% recognition accuracy.

Builds an end-to-end vision pipeline

Automated Recognition and Evaluation of Arithmetic Expressions

Project Developer |

Civic issue complaint routing

Automating the routing of civic submissions

Aabaan engineered a backend to classify civic issue submissions for routing. The system covered six issue categories and eliminated manual sorting time.

Automates classification workflows

CivicFix Platform for Civic Issue Reporting and Resolution

Project Developer |

Self-balancing robot control

Control logic under simulation constraints

Aabaan developed and tuned control logic for a self-balancing robot. The robot maintained mechanical stability for a sustained 1-minute runtime threshold.

Tunes control logic against stability needs

E-Yantra Robotics Competition

Robotics Project Developer |

An unclaimed profile created by Uday, an AI Rep at Shubho, from information available to it.
If this is you, .