Projects

A comprehensive n8n workflow that automates job hunting by cross-referencing resumes with live JSearch listings and scoring them using Google Gemini AI.

A lightweight, zero-dependency ESM utility library for vector mathematics and AI embeddings — built for semantic search, RAG pipelines, and ML workflows in JavaScript and TypeScript.

A custom-built n8n workflow that leverages Google Gemini to automatically parse and extract job titles from resumes with high semantic accuracy.

Master Thesis project implementing a high-performance framework for multi-modal emotion recognition using advanced self-supervised representation learning on audio and visual streams.

COLMAN explores object goal navigation using embodied AI agents, leveraging Transformer-based architectures and CLIP semantic embeddings for improved scene understanding.

The IndoRE project focused on relation extraction for three low-resource Indian languages: Bengali, Telugu, and Hindi using deep generative models and transformer architectures.

Research focused on robot skill acquisition through imitation learning and VR teleoperation, mapping raw pixels to complex robotic manipulation actions.

Implementation of a personalized gaze estimation framework using DT-ED networks and meta-learning to achieve high accuracy with minimal calibration samples.