Prashant

Key Skills

Software

Don't disclose

Top Subject Matter

LLM evaluation and optimization

Medical image/clinical diagnosis (tabular medical data)

Computer vision deepfake detection

Top Data Types

Image

Top Task Types

Diagnosis

Classification

Freelancer Overview

LLM based response evaluation & optimization system (project). Core strengths include Don't disclose, Streamlit, and OpenCV. Education includes Master of Technology, Jamia Millia Islamia (2026) and Master of Science, Swami Vivekanand Subharti University (2022). AI-training focus includes data types such as Computer Code, Programming, and Medical and labeling workflows including Evaluation, Rating, and Diagnosis.

Labeling Experience

LLM based response evaluation & optimization system (project)

Don't disclose

Built an LLM pipeline for response evaluation and optimization using BERT embeddings, custom metrics, and prompt engineering. Focused on measuring response quality and improving generation behavior based on evaluation signals. Produced real-time analysis with response classification and performance visualization via a Streamlit app. • Used BERT embeddings • Applied custom evaluation metrics • Performed prompt engineering for optimization • Implemented real-time classification and visualization in Streamlit.

2024 - 2026

Deepfake Detection by Deep Learning (project)

ImageClassification

Developed a deepfake detection system using ResNet-50 with transfer learning in PyTorch. Implemented OpenCV-based face extraction and data augmentation to improve robustness for real-vs-fake classification. Optimized training with Adam optimizer and binary cross-entropy loss, evaluating performance with accuracy, precision, recall, F1-score, and ROC-AUC. • Used transfer learning with ResNet-50 • Applied OpenCV face extraction • Performed data augmentation • Trained and evaluated with standard classification metrics.

2023 - 2026

Heart Disease Detection using Machine Learning and Python (project)

Diagnosis

Developed a heart disease detection classification model to predict whether a person has heart disease. Conducted feature analysis and evaluated model recall to support early disease detection. Work involved preparing datasets and training/testing ML models for medical diagnosis classification tasks. • Built an XGBoost and Random Forest based classifier • Performed feature analysis • Evaluated recall for early detection • Deployed using Streamlit for predictions.

2020 - 2022

Education

Work History