Developed ML/NLP models (urgency sentiment analysis, neural time series price forecasting, competition classification using weak labelling and data augmentation techniques) to help clients identify procurement opportunities, managing and providing technical guidance (model/architecture selection, etc.) to a team of 4 interns.

Built a conversational agent for retrieval over structured and unstructured public elections data.

Built a semantic search engine using for efficient retrieval of public records data for data journalists.

Research on legislative affairs + scraping and visualization of data on environmental rulemakings to inform clients on historic trends in regulatory action.