Let people chat with your knowledge

You know a lot about your subject. The problem is that almost none of it is in one place. Some of it is in a notes app you never share. Some is in a folder of PDFs. Some is on your website, buried three clicks deep. And a surprising amount of it lives only in the emails and messages where you have answered the same question, in your own words, dozens of times.

When someone shows up with one specific question, none of that helps them. They cannot read your notes app, and you cannot be awake for every question that arrives. So they get a search box that returns links, or a contact form that returns silence, and the thing you actually know stays locked in files only you can open.

There is a simpler shape for this now. Instead of asking people to dig through your material, you put all of it behind one link and let them ask.

What you are actually making

The thing you make is a lens: one shareable link where anyone can ask a question and get an answer drawn only from your own material, with the sources shown. You type a question and get an answer grounded in your own material. When your material does not cover something, the lens says so honestly instead of guessing, and the question lands in your inbox with the person who asked it attached.

That last part matters, so it is worth being plain about it. A lens is not a generic bot reciting a script, and it is not a search box with a nicer font. It is your knowledge made askable. Everything it says comes from what you gave it, and every answer can be traced back to the source it came from.

This post is about making your knowledge answerable. If what you want is a version of you that speaks in your voice and carries your personality, that is a slightly different angle, and we cover the persona side in what a digital twin is. Here the star is the knowledge, not the persona.

What to feed it

A lens is only as good as what you put into it, and the good news is that you almost certainly already have enough. You are not writing anything new. You are pointing it at what exists.

The material that works best is the material you already reach for:

• Notes and written text. Paste in the explanations you have typed out before, the rough drafts, the internal docs, the things you know cold.
• Documents. The PDFs, guides, and reports that hold the detail. In the Mindola app you can upload these directly.
• Your website and other links. Add the pages that already explain your work, and the lens reads them for you.
• Past answers and FAQs. The questions you get over and over, paired with the answers you have already given. This is often the most valuable material you own, because it is written in the exact language people use to ask.

Behind the scenes, everything you add runs through the same pipeline: it gets read, split into passages, and indexed so the lens can find the right piece for a given question and answer from it with a citation. You do not touch any of that machinery. You provide the knowledge, and the retrieval, the grounding, and the sources come with it. The more you feed it, the more it can answer, so it is worth adding your real material rather than a thin summary.

You do not have to load everything on day one. Start with the material that answers the questions you get most, publish, and add the rest as you go. A lens with your five most common answers in it is already more useful than a contact form, and it only gets better from there.

What a visitor experiences

Picture someone landing on your link with a real question. Here is what happens next.

They can type, and the lens answers in plain language, grounded in your material, with the sources shown right there so they can see where the answer came from. They do not have to hunt through your site or read a whole PDF to find one line. They ask the question the way they would ask you, and they get a straight answer.

They can also talk. A lens can take a live voice call, so a visitor speaks their question out loud and hears the answer back, and they can interrupt mid-sentence the way people do in a real conversation. You choose whether the voice is cloned from a recording of yours or a natural stock voice. Either way the audio carries a machine-readable marking that identifies it as AI generated, so nobody is misled about who is speaking.

And when the lens does not know, it does not pretend. It tells the visitor honestly that the material does not cover their question, and it drops that question into your inbox with the visitor attached. So the moments your knowledge could not reach become a short list of people who raised their hand, waiting for you rather than lost. It works a little like a light CRM without you setting one up.

The whole point is that the visitor never has to know how any of this is built. They see a link, they ask, and they get an answer they can trust because it points back to your own words. What used to take an email and a two-day wait now takes one question.

Your knowledge stops being a pile of files only you can open. It becomes one answer, ready the moment someone asks.

How to make one

There are two ways to build a lens, and you can pick whichever fits how you already work.

From an AI assistant. If you have Claude, ChatGPT, Cursor, or another MCP client open, it can build the lens for you inside one conversation. You connect Mindola once, tell the assistant what your knowledge covers, and paste or point it at your material. It assembles a profile, you confirm it, and it hands back a share link that is live right away plus a claim link. You do not need an account to start. Open the claim link, sign up for free, and the lens is yours to keep. A lens you never claim simply expires after thirty days, so nothing lingers. The Claude connect guide walks through the setup.

In the app. If you would rather see everything on screen, you can build and shape the lens in the Mindola app. This is where you upload richer sources like PDFs, add links and Q&A pairs, set the tone, and decide what the lens should hand straight to you. The full walkthrough of how it fits together is on the features page, and the docs cover every path end to end.

Either way, you are not stuck with your first pass. You can keep adding material as your knowledge grows, and the lens gets better each time.

See one before you build

The fastest way to understand a lens is to ask one something. The Charles Darwin demo answers only from public-domain writing and admits what it does not know, so you can watch the grounding and the honest refusals for yourself. The gallery has a dozen more built the same way. Try a few, ask them hard questions, and notice where they stop. Then point one at what you know.

Frequently asked questions

What can I feed my lens?

Whatever already holds your knowledge: pasted notes and text, documents like PDFs, your website and other links, and your past answers shaped as questions and answers. In the app you can add all of these, including PDF uploads. Through an assistant or the developer API, you add text and links, and your assistant can pull the relevant content out of a document and add it as text for you.

Will it make things up about me?

No. A lens answers only from the material you gave it, and every answer shows its sources. When a question falls outside your material, it says so honestly and passes the question to you instead of inventing a plausible wrong answer. Answering from your own material, and being honest at the edges, is what makes it safe to put your name on.

Can visitors talk to it instead of typing?

Yes. A lens can take a live voice call, so a visitor asks out loud and hears the answer back, and they can interrupt it mid-sentence. You choose a cloned voice or a stock voice, and all generated audio carries a machine-readable marking that identifies it as AI.

Where do the questions it cannot answer go?

Straight to your inbox, with the visitor attached. So the questions your material did not cover are not lost. They become a list of interested people waiting for a reply, which is why a lens doubles as a light way to capture the folks who reached out.

Do I need an account to make one?

No. A connected AI assistant can create a lens for you with no signup. You get a live share link and a claim link, and you only sign up, for free, when you want to keep it. An unclaimed lens expires after thirty days.