Build a Living Research Hub with Perplexity Spaces and Internal Knowledge Search

Professionals doing recurring research, whether tracking a competitor, monitoring an industry, or managing a long-running client project, often lose time re-explaining context to an AI assistant in every new search. Perplexity's Spaces feature solves this by acting as a dedicated workspace where custom instructions, uploaded files, and conversation history persist, and the company has now extended this with Internal Knowledge Search, letting Pro and Enterprise Pro users search across both public web content and their own internal knowledge bases in the same query. To set up a Space, create it, give it a description, and define custom instructions describing exactly what you want tracked and how the answers should be formatted, for example asking Perplexity to track a company's news, pricing changes, and hiring activity and present findings chronologically with date stamps. You can then upload reference files (contracts, past reports, internal memos) directly into the Space, and with Internal Knowledge Search enabled, Perplexity can pull from these internal documents alongside live web results in a single synthesized answer. A concrete professional example: a competitive intelligence analyst creates a Space named after a competitor, sets custom instructions to flag any pricing, product, or leadership changes, and uploads the company's own internal competitor-tracking spreadsheet. Each week, the analyst simply revisits the Space and asks what has changed, and Perplexity synthesizes new public information with the historical internal notes, avoiding the need to manually cross-reference old research every time. This workflow is equally useful for project teams, where a Space can act as a central repository combining internal project documents with external market intelligence, giving every team member instant access to the same up-to-date context. Updating the custom instructions as the research focus evolves keeps the Space relevant over months, turning what used to be a series of disconnected searches into a single, continuously improving knowledge asset.