Sharon Irukan

Portfolio Website & Web Design 1. Rapter Technical Solutions: Website Design & Development Project Type: Podcast Website Role: UI/UX Design · Web Design · Front-End Development Client: Rapter Technical Solutions Platform: Web Tools: Figma, React / Astro.js, Tailwind CSS Live Link: https://www.rapter.africa/ Project Overview Designed and developed a modern, professional website for Rapter Technical Solutions, an integrated security and systems solutions provider. The project focused on presenting complex technical services in a clear, credible, and visually consistent digital experience. Objectives Strengthen brand presence and credibility Clearly communicate services and expertise Support client inquiries and lead generation Ensure a fast, secure, and mobile-friendly experience Scope UI/UX design aligned with brand guidelines Responsive website development with CMS SEO-ready structure and analytics integration Deployment and security configuration Outcome Portfolio 1 Delivered a polished, scalable website that positions Rapter Technical Solutions as a reliable and professional systems integrator, with a foundation ready for future growth. Portfolio 2 2. Geohabari Podcast — Website Design & Development Project Type: Podcast Website Role: UI/UX Design · Web Design · Front-End Development Client: Geohabari Podcast Platform: Web Tools: Figma, React / Next.js, Tailwind CSS Live Link: https://geohabari.com/ Project Overview Geohabari Podcast is a media platform focused on conversations around geospatial technologies, mapping, urban development, and African spatial stories. The goal was to design and build a modern, credible website that showcases podcast episodes, strengthens brand identity, and makes it easy for listeners to explore content across platforms. Design Approach The design emphasizes clarity, structure, and personality. A bold hero section introduces the podcast, while a clean layout highlights episodes, hosts, and themes. Typography and spacing were optimized for readability, and the overall visual language balances tech professionalism with media storytelling. Portfolio 3 Key Features Responsive, mobile-first layout Episode listing with clear call-to-action links Podcast branding integrated into UI elements Optimized performance and SEO-friendly structure Scalable design for future content and episodes Outcome The final website gives Geohabari Podcast a strong digital presence, improves content discoverability, and provides a solid foundation for growing the podcast audience. Portfolio 4 Machine Learning and Web Development Modelling Soil Organic Carbon Using Satellite Imagery and Machine Learning Project Type: Final Year Research Project Role: Data Scientist · GIS Analyst · Machine Learning Engineer Study Area: Sichei Area, Western Kenya Tools & Technologies: Python, Google Earth Engine, Sentinel-1 & Sentinel-2, Random Forest, XGBoost, SVR, QGIS Project Overview This project focused on modelling and predicting Soil Organic Carbon (SOC) using multi-source satellite imagery and machine learning techniques. SOC is a critical indicator of soil health, agricultural productivity, and carbon sequestration, yet traditional field-based measurement methods are costly and spatially limited. Approach Field-sampled SOC point data were integrated with Sentinel-1 SAR and Sentinel-2 optical imagery to extract relevant spectral and radar-based predictors. Multiple machine learning models were developed and Portfolio 5 compared, including Random Forest, XGBoost, and Support Vector Regression (SVR), to evaluate their performance in capturing spatial patterns of SOC. Key Highlights Integration of SAR and optical satellite data for environmental modelling Feature extraction and preprocessing using Google Earth Engine Comparative evaluation of multiple machine learning algorithms Spatial prediction of SOC across the study area Best model achieved an R² of 0.54, demonstrating the viability of remote sensing–based SOC estimation Outcome The project demonstrates how remote sensing and machine learning can be combined to produce scalable, cost-effective soil carbon estimates. The resulting SOC maps provide valuable insights for land management, climate studies, and sustainable agriculture planning. GitHub Repository: GitHub repository Portfolio 6