32dots HEIDELBERG AI
Session 5 easy

Capstone: a sourced literature or market scan, end to end

LESSONLesson 5 · ~25 min

🎯Goal. Run one complete, verified scan — a literature scan for a research topic or a market/competitor scan for a venture — combining everything: a specific question, the right mode, citation verification, and a Space to hold it.

▶ Try this prompt

Researcher: "Give me a structured scan of the last 3 years of literature on lipid nanoparticle delivery for mRNA therapeutics — group by approach, cite peer-reviewed sources, and note open questions." Founder: "Scan the current market for at-home microbiome testing kits — list the main companies, their pricing, their claimed science, and link a source for each."

Pick the prompt that fits you. Run it in Academic + Pro Search, then verify the load-bearing claims by opening their sources — this is the full workflow, not a single query.

Steps
  1. 1Set up. Open (or create) a Space for the topic so the whole scan stays in one place and you can keep refining it later.
  2. 2Run it in the right mode. Academic source + Pro Search for a literature scan; default web + Pro Search for a market scan. Ask for a structured output — grouped sections or a table with a source per row.
  3. 3Drill with follow-ups. "Now only in-vivo studies", "add each company's funding stage", "what's the strongest counter-evidence?" — sharpen the scan without restarting.
  4. 4Verify the load-bearing claims. Open the citations behind the numbers and conclusions you'd actually act on. Mark anything you couldn't confirm as unverified rather than dropping it silently.
  5. 5Go deeper if you have Pro/Max: Perplexity's deeper research mode (and Labs, on paid plans) can run a longer multi-step investigation and assemble a fuller report or assets — useful for a first full draft of a review or a market map. Treat its output the same way: verify before you trust.

You'll see. A structured, grouped scan — by research approach or by competitor — where each claim carries a source, the weak spots are flagged, and the whole thing lives in a Space you can return to and share.

💳Cost. The scan and verification work on the free tier (within Pro Search limits). Spaces, higher limits, and Labs-style deep reports are Pro/Max features.

💡Takeaway. A real scan is a workflow, not a question: specific ask → right mode → structured output → follow-ups → verify the sources that matter → keep it in a Space. That's Perplexity used the way it's meant to be.

How was this lesson?