The Scenario
You're a researcher at a financial services firm. A colleague left last month, and you've inherited their workflow — part of which involved maintaining a background reading sheet on regulatory topics for the compliance team.
The sheet is supposed to contain the most relevant Wikipedia reference pages on whatever topic compliance is currently focused on. This quarter it's "SPAC merger regulation and investor protections." The sheet is empty. Your colleague never built it.
You could search Wikipedia manually. But the topic is technical enough that keyword search returns a mix of directly relevant pages and loosely related ones. What you actually need is semantic relevance — pages that are substantively about SPACs, merger regulation, and investor protection frameworks, not just pages that mention those words.
The bad version:
- Search Wikipedia for "SPAC merger." Open the top result. Decide whether it's actually about regulation. Copy the title and URL. Paste into the sheet.
- Search "investor protections SPACs." Skim results. Copy another three. Realize two overlap with the first search.
- Deduplicate. Reorder. Format the two columns consistently. You've got 12 pages but you're not confident they're the most relevant 12.
You need 20 pages. Confident ones.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent that lives inside your Google Sheet. Through its Perigon integration, it can run a vector search specifically against Wikipedia and write the semantically closest pages — title, URL, and snippet — directly into your sheet.
Use Perigon vector search on Wikipedia for the query "SPAC merger regulation and investor protections" and write the top 20 results to the Wikipedia Refs sheet with title, URL, and a one-line summary.
What You Get
- 20 rows in the Wikipedia Refs sheet, each representing one semantically relevant Wikipedia page.
- Column A: page title. Column B: URL. Column C: one-line summary or snippet.
- Ranked by semantic relevance to your query — not by keyword overlap.
- Ready to share with the compliance team immediately.
What If the Data Is Not Quite Ready
The query should be read from a cell
Read the research query from cell B1 of the Config sheet. Use Perigon vector search on Wikipedia for that query and write the top 20 results to the Wikipedia Refs sheet with title, URL, and a one-line summary.
You want fewer results but higher confidence
Use Perigon vector search on Wikipedia for "SPAC merger regulation and investor protections". Return the top 10 results only. Write title, URL, and summary to the Wikipedia Refs sheet.
You need to cross-reference against pages already in a previous reading list
Use Perigon vector search on Wikipedia for "SPAC merger regulation and investor protections". Get the top 20 results. For each page, check if the URL already exists in column B of the Previous Reading List sheet. Write title, URL, and summary to the Wikipedia Refs sheet. In column D, mark "new" or "repeat" based on the check.
Pull semantic Wikipedia results, categorize by subtopic, and build a structured index in one pass
Use Perigon vector search on Wikipedia for "SPAC merger regulation and investor protections". Retrieve the top 20 results. Write title, URL, and summary to the Wikipedia Refs sheet. In column D, assign each page to one of three subtopic buckets — "SPAC structure", "merger process", or "investor protections" — based on the title and summary. Count pages per bucket and write the totals to the Summary sheet.
Ask for the search, categorization, and summary in one prompt.
Try It
Get the 7-day free trial of SheetXAI and open any Google Sheet where you're building a reference library or background reading index, then ask it to run a Perigon Wikipedia semantic search and populate the results. See the full Perigon integration overview or explore building a one-shot research brief using Perigon and Wikipedia.