Transform months of manual data entry into hours. Extract entities, structure events, and build quantitative datasets from qualitative sources— powered by LLMs and AI Agents, and designed for researchers.
Automate the manual work of extracting data from hundreds of documents. Focus on analysis, not data entry.
Define and extract custom attributes and variables from unstructured text sources using LLM-powered parsing for flexible data structuring.
Collect data from news articles, reports, academic databases, and policy documents in parallel.
Full methodology transparency and reproducible workflows that meet academic research requirements.
Convert qualitative text into quantitative datasets ready for statistical analysis.
Built-in quality checks and validation ensure data accuracy and consistency.
Reach out to discuss your research needs
A full-stack ML system that predicts likely security responses to protests in Iraq, Lebanon, and Egypt. Trained on 13,000+ historical protest events, the ensemble model forecasts outcomes including verbal coercion, detention, and physical force — served via a REST API with a live interactive frontend.
Production system with 83-test CI/CD pipeline, Prometheus + Grafana monitoring, and separate backend API hosted on Render.
Please feel free to get in touch to discuss potential opportunities or view my work in detail.