Samwel Wekesa

Key Skills

Software

Kili Technology

Label Studio

LightTag

Roboflow

SuperAnnotate

Toloka

Top Subject Matter

No subject matter listed

Top Data Types

3D Sensor

Document

Geospatial Tiled Imagery

Top Task Types

Classification

Relationship

Translation Localization

Freelancer Overview

As a marine engineer with a strong background in complex systems analysis, I bring domain expertise in sensor data interpretation, technical documentation, and engineering schematics—skills directly applicable to AI training and data labeling. I have hands-on experience working with engine room systems, propulsion units, and onboard diagnostics, which enables me to accurately label time series data, identify anomalies, and classify technical events in sensor outputs. My familiarity with structured maintenance logs, system manuals, and safety procedures positions me well for NLP-related tasks such as entity recognition, document classification, and text summarization. I also have a working knowledge of annotation tools such as Label Studio, CVAT, and MATLAB for sensor data analysis. My ability to bridge engineering knowledge with precise labeling techniques ensures high-quality, context-aware AI datasets for machine learning applications in industrial, maritime, and technical domains.

Labeling Experience

Sensor Data Annotation for Marine Engine Fault Classification

Label Studio3D SensorEntity Ner ClassificationClassification

This project involved annotating sensor datasets collected from marine propulsion and auxiliary systems, focusing on engine temperature, vibration, and pressure logs. My role was to classify normal vs. abnormal behavior, label specific fault events, and tag temporal sequences for predictive maintenance model training. Additionally, I worked on summarizing maintenance logs using NLP techniques to extract operational patterns and recurrent issues. Quality control involved cross-validation with engineering standards, and labels were reviewed against known historical failures to ensure accuracy. The annotated dataset was used to train a diagnostic model for early fault detection in shipboard engine systems.

This project involved annotating sensor datasets collected from marine propulsion and auxiliary systems, focusing on engine temperature, vibration, and pressure logs. My role was to classify normal vs. abnormal behavior, label specific fault events, and tag temporal sequences for predictive maintenance model training. Additionally, I worked on summarizing maintenance logs using NLP techniques to extract operational patterns and recurrent issues. Quality control involved cross-validation with engineering standards, and labels were reviewed against known historical failures to ensure accuracy. The annotated dataset was used to train a diagnostic model for early fault detection in shipboard engine systems.

2022 - 2022

Education

Work History