ABDUL BASITH
Data Scientist & ML Engineer.
I build predictive analytics systems that help teams make faster, smarter decisions from data.
My work combines Python, Scikit-learn, SQL, and MLOps to move ideas from analysis into reliable production workflows.
I’m especially interested in public sector impact, where data-driven models can improve planning, monitoring, and outcomes.
My work spans full stack development, machine learning, and building data-driven products.
Design and ship responsive web applications with clean UI, good performance, and strong developer experience.
Create reliable APIs and backend services for real products, including authentication, integrations, and data workflows.
Develop machine learning models and RAG systems to turn data into practical, production-ready features.
Build dashboards and analytics workflows that support better decisions and measurable outcomes.
A selection of full stack and AI projects.
Automates financial risk analysis, anomaly detection, and continuous controls monitoring using machine learning.
Built an intelligent audit analytics platform to streamline audit workflows with automated anomaly detection and risk insights.
Answers questions from PDFs completely offline using local LLMs and semantic search.
Implemented a Retrieval-Augmented Generation pipeline with local inference, embedding-based retrieval, and a React UI for interactive Q&A.
Governance workflows for monitoring, compliance checks, risk assessment, and explainability for AI systems.
Built an end-to-end governance platform to improve transparency and accountability through automated checks, workflows, and reporting.
Predictive analytics platform that identifies customers likely to churn using machine learning and interactive dashboards.
Designed an end-to-end churn prediction workflow with feature engineering and a dashboard to explore key churn drivers.
Machine learning model to predict temperature variations from environmental pollution indicators.
Built regression models and evaluation workflows to understand how pollution indicators correlate with temperature variation.
NLP-based analysis of customer sentiment to generate actionable business insights from reviews and social media.
Built an NLP pipeline to classify sentiment and surface key themes that inform business decisions.
Worked on data-driven solutions to improve public service analytics and governance using machine learning and data science.
• Built ML models with Python, Scikit-learn, Pandas, and SQL on citizen grievance datasets.
• Performed preprocessing, feature engineering, and predictive modeling to generate actionable insights.
• Built dashboards and analytics workflows for visualization and decision support.
Built responsive web applications and collaborated closely with backend developers and designers.
• Developed React applications and integrated REST APIs.
• Improved page load performance and reduced UI latency through optimization.
• Built reusable components and improved cross-device compatibility.
Technologies I use across full stack development, machine learning, and product building.