Bhuwan Awasthi

ML Engineer
$347 / month
November 25, 2002

About Candidate

I am a Machine Learning Engineer with hands-on experience in designing and deploying end-to-end AI solutions. Proficient in Python, SQL, TensorFlow, and PyTorch, I have worked extensively on NLP, computer vision, and LLM-based projects. I also successfully sold an ML-based SaaS solution to a US-based firm, showcasing my ability to deliver impactful and scalable AI solutions.

Location

Education

B
B.Tech 2024
Madan Mohan Malaviya University of Technology

Computer Science and Engineering

Work & Experience

M
ML Engineer/Instructor Intern March 2024 - August 2024
National Institute of Electronics & Information Technology, Gorakhpur

Served as a Machine Learning, Data Science, and Artificial Intelligence instructor for certification courses run by AICTE under the Ministry of Electronics and Information Technology (MeitY), teaching Python, SQL, Machine Learning, Statistics, Probability, Deep Learning, AI, and mentoring over 500 students.Developed a ML model using ensemble learning, transfer learning, and CNNs to detect real and spoof faces during attendance on government education platforms.Addressed an imbalanced dataset (86% Spoof, 14% Real) by implementing class-weighted loss functions, oversampling, and advanced data augmentation, achieving 87.72% accuracy. Developed efficient data preprocessing pipelines for faster model training and improved performance.

S
Software Engineer October 2023 - December 2024
Adroit Market Research

Developed a market research automation solution using Python, leveraging NLP techniques, Selenium for web scraping to generate detailed market reports, boosting content creation efficiency by 50% and reducing manual work requirements by 25%.Automated data analysis and document processing with Pandas, NumPy, and win32com, packaged applications using PyInstaller, and reduced data processing time by 40%.

D
Data Analyst VPatrol AI, Gurugram - January 2025

Designed and implemented an advanced anomaly detection system using Isolation Forest and Z-Score normalization, achieving a 30% improvement in multi-dimensional outlier detection accuracy.Automated high-dimensional data augmentation workflows with NumPy, Pandas, and Scikit-learn, integrating custom feature engineering and hyperparameter tuning, reducing preprocessing latency by 40%.