M Thatcher Clark describes his work – Pacer Tracker – available on Internet Archives – “At its creation in April 2024, this database contained the metadata of more than 350 million docket entries filed since 2013 in more than 13 million federal court cases. The data was obtained from RSS feeds published by many of the federal courts. This includes many bankruptcy, district and appeals courts. It also includes the Court of Federal Claims and the Court of International Trade. The data is not comprehensive, as some courts do not publish a feed and some of them do not include all filings in their feeds. A breakdown of how many cases and entries were available in the database at the time of its creation is contained in the entries_count.csv file. For each case, the data supplies the case’s court, its number, its title (including the primary parties), its PACER docket report website URL, the type of case, and the time when it was first captured, which is often roughly the same as it when it was filed. For each entry, the data supplies the entry description as it appeared on the docket, any PACER document URL for the entry, the time it was filed, the time it was captured and its docket sequence number (if any). Document URL and docket number data is often not available between 2018 and 2023. The database consists of a courts file, a cases file and entries files for each year. Each case has a court ID linking it to a court and each entry has a case ID linking it to a case. All three types of files are needed to fully comprehend the data. A dictionary containing more information about each field of the data files is contained in data_dictionary.csv. Statements suitable for PostgreSQL, which can be easily modified to any SQL-compliant database, are available in sql_import_stmts.sql. Updates to the data will occur at least weekly, to the courts, cases and most recent entries file.”