Open to opportunities

Banesori Ayekpam

@banesoriayekpam

Message

Machine learning and computer vision researcher building multimodal, temporal models that drive decisions.

India

Message

What I'm looking for

I’m looking for a role where I can build reliable ML/DL systems—especially multimodal, computer-vision, and temporal modeling—and turn complex data into decision-ready insights with real-world impact.

I’m a machine learning and computer-vision focused researcher with an M.Tech in Computer Science and Engineering from NIT Manipur. My work centers on reliable modeling across time and modalities—ranging from early radiographic progression to ransomware detection using ensemble learning.

In industry, I interned as a Data Scientist at 360DigiTMG, cleaning 2500 machine records and deploying ML models with 98.6% best accuracy via Streamlit, while reducing machine downtimes by 10% and cutting unplanned downtime with at least $1M in cost savings. I also delivered impactful analytics projects (data quality +50%, decision efficiency +35% through Power BI) and achieved strong research results, including water-bodies regression with R-squared 91% and ransomware detection accuracy of 80.08% using Random Forest.

Experience

Work history, roles, and key accomplishments

Current

Assistant Professor

Current

S-VYASA University

Mar 2025 - Present (1 year)

Delivered ~16 hours/week of undergraduate instruction in problem solving using Python, data science, and machine learning while designing course curriculum and learning materials.

Python Data Science Machine Learning Curriculum Design Teaching Jupyter Problem Solving

Ransomware Detection Research

National Institute of Technology Manipur

Jan 2022 - Jan 2023 (1 year)

Built an ensemble ML-based ransomware detection system on the CICAndMal2017 dataset (500K+ samples) using feature engineering and comparative algorithm analysis. Achieved 80.08% accuracy with a Random Forest model using a 25-feature subset selected via rigorous feature selection methods.

Ensemble Analysis Random Forest Feature Model Selection Machine Learning scikit learn Data Analysis Python

Research Intern - Water Quality

National Chung Cheng University

Analyzed 4,000+ aquatic images and paired water-parameter records to predict critical environmental parameters, achieving 91% R-squared using ML/DL models that combine CNN image features with tabular models. Developed a multi-modal data augmentation method that increased the available dataset by 4,200 records to address data scarcity.

Computer Vision CNN XGBoost Random Forest Regression Modeling Machine Learning

Data Scientist Intern

360DigiTMG

Cleaned and transformed 2,500 machine records using Python and SQL, enabling downstream analytics and decision-making. Built and deployed ML models with 98.6% best accuracy using Streamlit and monitored performance with evidently.ai, contributing to 10% reduced machine downtime and at least $1M cost savings.

Python SQL Data Cleaning Machine Learning Streamlit Evidently AI Matplotlib SeaBorn

Education

Degrees, certifications, and relevant coursework

National Institute of Technology, Manipur

Master of Technology, Computer Science and Engineering

2021 - 2023

Grade: CGPA: 8.25/10

M.Tech in Computer Science and Engineering at NIT Manipur (2021–2023); CGPA 8.25/10. Thesis: A Machine Learning Approach to Ransomware Detection Using CICAndMal2017 Dataset.