About

I am a multidisciplinary problem-solver who thrives at the intersection of technology, design, and analytics. My expertise spans a wide array of domains, from machine learning to interactive dashboards, with a particular focus on creating human-centered solutions. Whether it’s building predictive models for healthcare or analyzing data for business insights, I balance technical precision with user-centric design. I have a knack for handling class imbalances, optimizing models, and visualizing data, all while maintaining a meticulous approach to project organization. I embrace challenges like developing sentiment analysis tools, malware detection systems, and disease risk predictors, often leveraging tools like Streamlit, Tableau, and Python to turn raw data into actionable insights. With a commitment to precision and a love for innovation, I always looking to push boundaries and create impactful, well-thought-out solutions.

Skills

  • Data Science
    10
  • Data Analysis and Machine learning
    9
  • Data Analyst using Power BI and Python
    9

Experience

Kevin Muindi

Work experience
  • I am a highly skilled professional whose responsibilities encompass a wide range of technical and analytical domains. I excel in project management, leading the end-to-end development of data-driven applications, from problem identification to deployment. My work demonstrates a meticulous approach to handling large datasets, focusing on preprocessing, exploratory data analysis (EDA), and building predictive models. I consistently deliver impactful solutions tailored to meet technical, user-centric, and organizational goals, ensuring accessibility and usability through human-centered design. Additionally, I specialize in creating interactive dashboards and visualizations using tools like Tableau and Streamlit, enabling stakeholders to interpret complex data insights and make informed decisions.
  • My accomplishments reflect a strong track record of innovation and problem-solving. Notably, I developed the NHS Early Risk Detection Framework, a cardiovascular module aligned with NHS standards, which aids in identifying high-risk patients for proactive healthcare interventions. In fraud detection, I built a robust system leveraging PCA-transformed datasets, optimizing precision-recall metrics to handle highly imbalanced data. My Tableau-based Digital Business Intelligence Dashboard for a service organization exemplifies your ability to generate actionable insights through descriptive, predictive, and prescriptive analytics.
  • In the field of sentiment analysis, I created a multi-model system for classifying social media sentiment, achieving exceptional results with Logistic Regression, and deployed a Streamlit app for real-time sentiment classification. Furthermore, I designed an Android Malware Detection System that effectively identifies malicious applications using permission data, integrating advanced EDA and preprocessing techniques. Another notable achievement is my R-based predictive model for heart attack risk, utilizing Random Forest to support proactive healthcare planning. Across my projects, I've shown expertise in addressing challenges like class imbalances, improving model reliability, and ensuring fair and accurate predictions.
  • My ability to merge technical expertise, analytical rigor, and user-focused design makes me an exceptional problem-solver, consistently delivering innovative solutions that create meaningful impact across various domains.

Related persons

samuel kyamaHTML/CSS , Python, Javascript
joe kabucho