Skill · Perplexity Help Center & Beginners in AI
Build AI-Powered Research Workflows with Perplexity
Photo: Jo Lin on Unsplash
Perplexity Projects let you combine internal documents with real-time web research in a single persistent workspace. By setting custom instructions once, you create a reusable template that handles recurring research tasks - competitive analysis, market trends, regulatory updates - without re-prompting every time.
Perplexity Projects (formerly called Spaces) solve the recurring research bottleneck that many teams face: running the same research question monthly, only to start from scratch each time. A Project is a persistent workspace where you upload internal documents, set custom AI instructions once, and run all future research through those same rules.
Start by creating a Project for a research task you run regularly - quarterly competitive analysis, weekly industry news synthesis, or monthly regulatory compliance tracking. Upload 5-10 core reference documents: past reports, brand guidelines, competitor profiles, or domain context. Then write a 3-4 sentence custom instruction that specifies your research objective, preferred source types (academic journals, analyst reports, news), and output format (bullet points, comparison tables, executive summary). Example: "I am a market analyst researching enterprise SaaS in EMEA. Prioritize analyst reports from Gartner and Forrester published in 2025-2026. Return findings as a comparison table with pricing, GTM strategy, and customer concentration."
Once saved, every search you run inside that Project automatically anchors to those instructions and documents. You ask "What has Competitor X announced this quarter?" and Perplexity returns findings that reference both your internal history and current web information - a synthesis that would take you 30-45 minutes manually but happens in 3-4 minutes. The power is that the template persists. The next month, you ask the same question in the same Project and the instructions apply automatically.
The practical result: teams report 60-70% time savings on recurring research. A competitive intelligence manager runs the same analysis weekly but each run now takes 5 minutes instead of 30 minutes because the Project eliminates re-prompting and the custom instructions prevent context drift. Multiply that across a quarter and it compounds fast.
Sources & credits
Original source: Perplexity Help Center & Beginners in AI