Bartik provides practical implementations in AI, data analysis, OCR and scraping. We have developed tools in these areas that are currently in use by ~25 research groups worldwide. Relatedly we have worked on implementing related IT solutions for about 50 companies as of 2024.
We are based in Cambridge / Boston (Massachusetts, USA) where we have over 10 years of (scientific) experience working at Harvard University and related firms in the area. So far we have worked with firms, NGOs and research groups in the USA, Europe and Asia.
We have mainly worked on developing and implementing (open-source) Large Language Models (LLM) and related GPT models for business needs (e.g. to provide document analysis needs for companies, chatbots for internal use or customer support, or general natural language processing (NLP) tasks).
We have over 30 years of cumulative experience in working with some of the largest datasets in the world (e.g. Census data in different countries covering billions of records). We mainly work in Python, R and SQL and have related experience in STATA.
We built systems to digitize and analyze thousands of company invoices / consumer receipts as well as millions of historical documents and books.
We have developed scrapers that (1) gather policy proposals from thousands of government websites and (2) collect records on companies, job openings and other career-related data from millions of websites worldwide.
We collectively published 23 articles in scientific journals in the areas above and contributed to a number of related open source projects.
If you wish to work with us or have other enquiries, please click here to contact us.