Software that handles these steps is called e-discovery software.

The steps of e-discovery: preserve the data under a legal hold, collect it, process it so it can be searched, review it for what is relevant and what is privileged, and produce the relevant documents for the other side. At each step there is less data.

Why review costs so much. Review is the expensive step. Traditionally, lawyers read the documents one by one and are paid by the hour. It is among the largest costs of a lawsuit in America. Software has helped for years. Lawyers review a sample, and machine learning uses their decisions to rank the rest by how likely each document is to matter. The industry calls this technology-assisted review. American courts have accepted it since 2012.

Technology-assisted review: lawyers review a sample of the documents, machine learning learns from their decisions, and the rest are ranked by how likely each document is to matter.

Where the market is. Discovery is far broader in American courts than in India or most other countries. So most e-discovery software is sold in America, to law firms and to companies' legal departments. Much of the human review has been done by legal-services teams in India. This business is called legal process outsourcing.

Every company named here has posted software engineering jobs in India. Famous companies that don't actively hire software engineers in India are left out.