Dharmendra P.

Key Skills

Software

Other

Top Subject Matter

Deep Learning, LLM Inference and Training

Machine Learning

Graph Neural Networks and C/C++ Programming

Top Data Types

Text

Computer Code Programming

Top Task Types

Entity (NER) Classification

Freelancer Overview

Have experience of AI training on outlier.ai projects. Worked on various domains mainly programming, Full Stack web development, Machine Learning, Deep Learning. Research Intern — Explainable adaptive GNN-based financial fraud detection. Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other, Internal, and Proprietary Tooling. Education includes Master of Technology (Research), Delhi Technological University (2025) and Bachelor of Technology, Dr APJ Abdul Kalam Technical University (2024). AI-training focus includes data types such as Computer Code, Programming, and Text and labeling workflows including Evaluation, Rating, and Entity (NER) Classification.

Labeling Experience

Teaching Assistant — Machine Learning and core CS lab/tutorial support

TextEntity (NER) Classification

Served as a Teaching Assistant supporting instruction and guidance for undergraduate coursework in Machine Learning and core CS subjects. Provided mentoring that included assisting students with assignments, debugging, and academic projects that likely involved supervised learning and model training activities. Supported lab/tutorial delivery to help learners apply ML concepts during practical sessions. • Assisted faculty in conducting machine learning labs and tutorials • Mentored students on assignments, debugging, and academic projects • Helped coordinate learning support for OS and core CS lab/tutorials • Reinforced ML concepts during student practical work

2025 - Present

Research Intern — Explainable adaptive GNN-based financial fraud detection

Other

Worked as a Research Intern to build and evaluate a graph neural network for explainable, adaptive financial fraud detection on transaction graphs. Focused on improving model performance metrics (F1 score and AUC-PR) by handling class imbalance and concept drift in the training/evaluation pipeline. Involved experimentation with graph-based hybrid architectures (GCN, GATs) and state-space models to advance detection quality and robustness. • Developed and iterated model training/evaluation workflows for fraud detection in complex transaction graphs • Addressed class imbalance and concept drift to improve F1 score and AUC-PR over baselines • Researched explainable and adaptive GNN architecture approaches • Experimented with non-linear deep learning and graph-based hybrid models including state space models

2025 - Present

Education

Work History