Learn how to use the Findest Universe

From your first question to a project you keep returning to.

Nine short steps in 15 minutes.

AI for R&D 9 steps One step closer to your innovation

Ask your first question.

Guided Research is the main way to ask a research question on your Findest Universe. You type a question, and IGOR^AI searches for sources, analyzes them, and comes back with answers and citations to every paper it drew from. We'll show how it works under the hood in step 3.

What you do

Type a question → wait a couple of seconds → read the answer → click any citation to open the source.

What you get

A Quick Answer with clickable citations — plus three other views of the same papers, covered next.

Time

About 2–3 minutes per question.

Examples of good first questions

Be specific. Be answerable. Start with the question that's actually on your desk this week.

  • "What materials improve corrosion resistance of polycarbonate?"
  • "Which compression technologies are used for hydrogen storage above 700 bar?"
See the click-through Run a Guided Research, end to end

Expect to see the sources in about 1 to 2 minutes. While you wait, Findest is reading and judging roughly 3,000 abstracts in the background.

You can come back to your results any time — they're saved.

Try it now

Open Guided Research and ask the question that's actually on your desk this week. Read the Quick Answer when it lands.

Explore the results from four angles.

One Guided Research, four tabs. Same papers, different lens on each — switching tabs doesn't re-run anything. Each lens answers a different question.

Quick Answer

"What's the short version?" A paragraph synthesised across the top 30 papers, with numbered citations to every source.

Results Table

"What are the solutions discussed in the documents? How do they compare?" Technologies in rows, extracted attributes in columns. Click > on any row to see its papers.

Grouped Results

"What solutions are in my results? Which papers are linked to them?" Lets you navigate to the documents behind each group of results.

Top Sources

"Which papers should I read first?" Highest-relevance papers ranked, with information extracted, year, authors, institution, open-access flag.

See where each lens lives The four views in the left sidebar of Guided Research

The Results Table — Findest is extracting structured fields from each paper. Quick Answer, Grouped Results and Top Sources load instantly.

On the Results Table, the small > at the start of each row expands to show detailed information extracted from the documents of that group.

Why you can trust the result.

A Guided Research runs a multi-step dedicated researcher loop against 263 million scientific works in our database. It is based on our experience scouting for over 3000 projects for R&D teams. Every sentence in your answer is tied to a specific paper. If IGOR can't find an answer, it tells you that, instead of inventing one.

5 search variations per question
3,000 abstracts read per search
7–8 iterations until results stop changing by changing keywords or variables
263M scientific works in the corpus

When IGOR is wrong, it's wrong in a boring way. The failure mode is a relevant paper didn't make the top results — not a fabricated answer. Every sentence in the Quick Answer is tied to a paper you can open. If the source doesn't back the claim, you'll see it in one click.

What it won't do. Ask "how many canals are in Amsterdam?" and IGOR returns "no information found in the linked sources" if it does not find documents about canals — not a guess. It only answers from what it has actually found and read.

The Judge Model, step by step

IGOR runs the same four-step loop on every Guided Research. Each step adds one specific thing you don't already know.

01

Extract keywords

Your question becomes 5 variations — including multiple synonyms and adjacent terms a human searcher would forget.

02

Search the corpus

Each variation runs against 263M curated scientific works. About 3,000 abstracts get pulled per question.

03

Judge each abstract

IGOR reads abstracts one at a time and scores relevance compared to your question — the same way you would, just faster. Irrelevant papers are dropped.

04

Refine and repeat

The relevant papers are used to generate new relevant keyword variations. The loop runs 7–8 times. It stops once more than 100 documents are found, or 2 minutes have passed.

More on the corpus and how to verify

Findest indexes new scientific works each week — think journal papers, conference proceedings, etc. 60 million of those are Open Access; once they're added to your project, the full text gets indexed. The corpus refreshes every week.

For context: Scopus ≈ 97.3M Web of Science ≈ 92M Google Scholar ≈ 389M (uncurated)

How to verify a result. Click any citation or any cell in the Results Table — the original paper opens in the sidebar. If a claim doesn't match what's in the source, you would see it directly.

Share your results.

You can share what you found directly with colleagues, straight from your Guided Research results. Copy and drop the Quick Answer straight into a slide deck, or send a colleague a link to what you found. Both routes keep the citations clickable — so whoever reads it can check the sources themselves.

Copy with formatting

Click the copy icon → paste into Word, PowerPoint, email. Citations paste as live links — your reader can open any source without leaving the doc.

Share via invite

Invite by email. The recipient gets a link to the live results view — they don't re-run the search, and the page stays current. Access is restricted to the addresses you list.

No project required

Both work directly from your results. You can also save and modify them into a project in step 5 if you want to keep building.

See the click-through Copy with citations, or share via invite

Sharing a result is not the same as making your project public. Share gives view access only to the specific people and results page you choose — projects stay personal (more on that in step 5).

Save what you found into a project.

If you copied or shared from step 4, that was the one-off path. This is the keep-building path. A project is a container for one research topic — it holds all your questions, IGOR's answers, the sources you saved, and the pages you build from them. You'll shape those pages in step 7.

Where it lands

Quick Answer → saved as a page in the project sidebar. Results Table → saved as a tab in the Project Overview. Each Grouped Results theme → its own page.

Visibility

Projects are personal. Only you see yours. Use Share (step 4) to send specific results outside.

To create

About 30 seconds. Name and short description — that's it.

What a project looks like once you save Opening a populated example — yours starts empty and fills as you save into it

The 💾 icon on Quick Answer and the 💾 icon on Results Table save to the same project but land in different places.

Quick Answer → sidebar page. Results Table → Overview tab. This is intentional, but easy to miss the first time.

Ask IGOR a follow-up question, or get help building your report.

Ask IGOR will help you to ask your own questions to the sources and speed up your work. The sources you saved in step 5 are what IGOR reads on each project page. Ask follow-up questions in plain English against your project's evidence — Ask IGOR returns a cited answer you can verify in one click, and turn into a page in another.

How do you find Ask IGOR? The ⚡ icon — anywhere you see it in your project — takes you to Ask IGOR.

Q&A mode

Ask any question and IGOR^AI scans the documents, looks for clues, combines the answers, and gives you a response. You can ask a simple question, draft a section of your report, and more.

Extract Detail mode

When you want information specific to each document, IGOR^AI reads each document one by one and fills a structured table. Best for comparison, verification, and large source sets.

Also available: re-built prompts for the most common research tasks
Summarize

Quick paragraph across all linked sources. Start here.

Write section

Drafts a report section with inline citations — ready to paste into a doc.

General description

Neutral overview of a topic, no comparison or judgement.

Standard report

Long, structured write-up across multiple aspects of the topic.

See the click-through Example of a quick answer

Shape your project: add pages, reorder them, edit them.

You just created pages — here's how to grow them. A page is the unit your project is built from. Saved searches arrive as pages. IGOR answers can become pages. You can write blank pages yourself. The five minutes you spend reordering and editing now is what makes the project still readable to you — or to a colleague — three months from now.

01 · Add

Click + in the project toolbar for a blank page. Or use Add to page on any IGOR answer (step 6).

02 · Order

Hover a page in the left sidebar — the drag handle appears. Drop it where it should live in the structure.

03 · Edit

Open the editor, rewrite anything, click Connect sources to bring more documents into the page's scope.

When to spend time re-ordering your project? If you are getting too many pages, or many pages end up being similar, it is a good time to start re-ordering your project to keep it simple to use in the long term.

Two ways pages come into being Saved from search (left) — or written from scratch (right)

Save from the web with the extension.

R&D evidence can also live outside the scientific or patent corpus — supplier pages, internal docs, news, vendor PDFs. You run into this kind of information in your project all the time, and you'll want to save it to a project and make it usable. With the Findest browser extension, you can save any webpage into a project as a source, so IGOR can read it next time you ask a follow-up.

How to install and use the extension?

The extension works for Chrome and Edge:

  1. Install the extension
  2. Open any page — PDF, supplier site, internal doc.
  3. Click the Findest icon → pick a page → it saves as a source.
  4. The source is now available in your project and for Ask IGOR to dig into!

Heads-up. Some sites (Espacenet, paywalled journals, intranets) block extraction — a work-around is to select and save the whole page as a highlight, or to upload the PDF to the relevant project page directly.

See the extension in action

Keep the project growing.

Projects expand on their own. Every finished page raises a follow-up — a new material, a new mechanism, an adjacent topic. The loop below is how you can keep one project useful for months.

The signal it's working: you start asking IGOR follow-ups instead of running new searches.

Three patterns customers fall into: a weekly question into one project (Monitoring), a monthly summary ask to IGOR (Meeting Prep), or a per-experiment debrief project (Lab Debrief). Pick one — don't pick three.

The growth loop, four moves

Pages stay searchable. Saved sources stay attached. Projects you started months ago remain useful when you return — which is the whole point.

01

Spot a follow-up

From a page you've built, identify a new question.

02

Run Guided Research

On the new question. Same flow as step 1.

03

Save to the project

Same project for follow-ups. New project for an adjacent topic — keeps things tidy.

04

Ask IGOR again

Integrate the new evidence. Use Add to page to keep the answer.

Anything you save from the web with the extension (step 8) counts as a follow-up too.

Now go find your answer.

Everything you just read is one click away in your Universe — ask a question, save what matters, and let the project grow as your research does. Keep this page bookmarked; it's your reference whenever you need it.

Open the Universe →