Guide to ScanEron

How ScanEron works

The complete overview in a minute of reading.

ScanEron turns a sticker with a QR code into the always updated page of your product, software, or service. You upload the documents once; those who purchase or use it scan the code with their phone and find everything: downloadable documents, videos, features, frequently asked questions, and an assistant that responds in its own words by reading what you uploaded.

1. Create a QR: choose whether it describes a physical product, software, or a service, and give it a name and a description.
2. Upload the contents: documents, diagrams, photos, videos, features.
3. Press Save and process: ScanEron reads the documents and prepares the assistant.
4. Set the QR to Active and print the sticker with the QR.
5. The customer scans and finds everything. You see in Analytics how many times it has been scanned.

Eron, the assistant

Who is that hawk that appears in the corners

The ScanEron assistant has a name: it’s called Eron, which is the second half of Scan-Eron, and has the face of the brand's hawk, the one with the eye made like the reference square of a QR code.

You meet it in two different places, and it does two different jobs:

• Inside the site, for you. It’s the button at the bottom right on all pages: you ask it how to do something and it responds by looking at this guide. It doesn’t touch your data and doesn’t modify anything: it just explains.
• Inside the pages of your QR, for your customers. There it responds only with what you have uploaded — documents, sheets, instructions, frequently asked questions. If something is not in your documents, it says it doesn’t know: it doesn’t make things up and doesn’t go looking elsewhere.

Eron is the starting avatar of all the QR codes and all the chats. If you prefer another face, in the AI tab of the editor you can choose another avatar or upload your own — for example, the face of the person in the company who actually provides assistance.

Registration and email confirmation

Upon registration, you receive an email with a six-digit code. Enter it on your first access to fully activate the account.

Before confirmation, you can enter and view your account, but you cannot upload files or publish QR codes. If you do not confirm within 48 hours, the profile is blocked: to reactivate it, just log in and enter the code, even after the deadline. If the email doesn’t arrive, from the code page you can request a new one.

Create the first QR

1. Go to My QR and press New QR.
2. Write the name of the QR: it’s the title that the customer sees first.
3. Add brand or supplier, model or formula, and a brief description. They also help the assistant understand what it’s about.
4. Choose the language of the public page.
5. Save: the QR is created as a draft, visible only to you.

Each QR receives an address with a random code, for example www.scaneron.com/QR/s6H7gAD41ikjTF5. It’s that address that ends up in the QR: it never changes, even if you rename the QR or replace the documents.

In the Information tab, you also find the QR options:

• Track statistics — counts scans, downloads, and questions.
• AI chat active — shows or hides the assistant on the page.
• Document downloads — allows or prevents file downloads.
• Show FAQs — makes frequently asked questions appear.
• Visibility — public, password protected, or private (only you).
• Company logo — at the top above the name or at the bottom in Activity Data.

The guided creation

When you don't know where to start, start here.

Next to New QR is Guided New QR. It does the same thing, but one question at a time, and at each step the assistant explains what it's asking you for. There are six steps and you can go back and forth freely.

1. The QR — name, type, category, and language. This is where the draft is created: from this moment the QR exists and its address is already the final one.
2. The description — two lines about what it is. If you don't know what to write, Write it for me: the assistant suggests one and you can correct it.
3. The documents — documentation and, for products and software, technical documentation.
4. The photos — those of the QR; the first one becomes the cover.
5. The ready-made answers — a button, and the assistant reads what you've uploaded and fills in features and frequently asked questions.
6. Finish — the summary of what's inside, and the choice: publish now or leave it as a draft.

If you skip a step, the assistant will tell you once what you're missing — without a document the chat has nothing to respond to, without photos the page shows the code instead of the QR — and on the second press of Next it lets you pass. The Skip this step button moves on immediately, no discussion.

Upload the contents

Everything the customer will find on the page, and that the assistant will use to respond.

Documents and technical documentation. Above the upload area, choose which of the two sections the files will go to: Documentation for documents, instructions, warranties, and certificates; Technical documentation for electrical schematics, CAD drawings, and exploded parts. A file can be moved from one section to another at any time with the icon next to its name.

You can drag files directly onto the box. These formats are read and indexed:

Displayed name. A document is often called blueEVO_S_plus_230V_ITALIA_8213_TS.pdf. With the pencil icon, you can show it as "Document Evo 230V": whoever downloads it still receives the file with its original name.

Video. Paste a YouTube or Vimeo link: the first YouTube video is shown directly on the page, without taking the customer out of the site.

Written instructions. Text that you write. For each block, choose whether to make it visible to the public or only for the AI: in the second case, it doesn't appear anywhere and is solely for instructing the assistant. To change the choice on already saved text, use the icon next to it in the list.

Features. Pairs "feature → value" groupable by section. They appear as a tab on the page and are always passed to the assistant, as they answer most questions about data and measurements. With Quick Entry, you can paste them all together:

Power:
Voltage: 230 V
Frequency: 50 Hz

Dimensions:
Width: 60 cm

A line that ends with a colon becomes the name of the section.

FAQ. Questions and answers that appear on the page and that the assistant uses as a primary source. They are the fastest tool to cover a recurring doubt.

Automatic filling

Features and frequently asked questions, taken from the document.

In the Features and Frequently Asked Questions tabs, next to Quick Entry, there is Automatic filling: the assistant re-reads the documents you uploaded and suggests the lines to add.

The suggestions come in a checklist, and they are fields, not text to look at: you correct them there, before saving. Nothing gets written until you press Add.

The question about language. Before starting, ScanEron checks what language the files are really in. If they are all in the language you are working in, it doesn’t ask anything; if it finds an Italian document while you are working in French — or two documents in two different languages — it stops and lets you choose which language you want the results in.

Check for errors in the document

ScanEron re-reads a document and tells you what is NOT written.

A document is written once and printed for ten years. The time to realize it doesn’t say where to dispose of the QR is before sending it to print, not when a customer asks for it.

In the Contents tab, next to each document, there is the shield icon. It opens the verification box: the assistant re-reads the document from the beginning and answers one question — what is missing?

It doesn’t check if the document is well written: it checks if it says the things a document must say. Safety warnings, manufacturer data, installation and maintenance, disposal, reference to the declaration of conformity.

The box states three things, in this order:

• The completeness score, in percentage — ‘complete document’, ‘to be completed’, or ‘important information missing’.
• What is missing — the gaps found, the most serious at the top, each marked as severe, medium, or minor.
• What is already there — because a list of only defects makes even a good document seem wrong.

The score stays attached to the document: in the file list, a tag with the percentage appears, so you can see at a glance which documents have already been checked and how they went. Reopening the box re-reads the previous verification and doesn’t cost a new one; to redo it, there is Redo the verification.

• It is done only on documents: there are no gaps to look for in a photo of an exploded view.
• The document must have already been read. If it hasn’t, the box says so: press Save and process in Contents and try again.
• With a very long document, the verification reads the first part of the document and writes it down for you.

The function is included in the higher plans: each verification is a complete reading of the document, and that’s why it’s not available in all plans. The shield, however, is there for everyone — with the lock if your plan doesn’t include it — and pressing it shows what it does, instead of discovering it by chance.

Protect downloads with a password

If certain documents must remain confidential to technicians, enable Password for downloads in the Contents tab. You can set two distinct passwords: one for documentation and one for technical documentation.

1. Enable the option and write the two passwords.
2. On each file, the lock icon rotates between three states: no password → documentation password → technical password.
3. On the QR page, protected files show 'Protected' and the box to unlock them appears.

Whoever enters a password unlocks only the corresponding group for the rest of the visit. The two passwords remain independent: you can give one to the end customer and the other only to the support centers.

Photo of the QR

In the Photo tab, you upload the images of the QR. The first one is the cover: it appears at the top of the public page and, in your dashboard, in place of the QR code.

• The other photos can be browsed with arrows, dots, or by swiping on your phone.
• By tapping the photo — or the lens icon in the top right — it opens in full screen.
• You can choose which photo is the cover and reorder them.

How many photos you can upload depends on the plan: 1 in the free plan, 5 in the Base, 20 in the Pro, unlimited in the Enterprise.

The AI assistant

How it responds, what it reads, and how to teach it what it doesn't know.

The assistant responds only with what you have uploaded: documents, instructions, features, and FAQs of that QR. It doesn't invent or draw from external sources. If the information is not there, it says so and invites you to contact support.

How to prepare. After uploading the documents, press Save and process. ScanEron extracts the text, divides it into indexed sections, and generates a summary. For scanned PDFs (those made with the copier, without selectable text), it uses the AI's visual reading, so it takes a few extra seconds.

Free text for the assistant. In the AI tab, there is a field where you can write what is not in the documents: internal procedures, known defects, error codes, support contacts, compatible spare parts. If the text is long, press Summarize this text: from that moment on, the summary is passed to the assistant, so the responses remain quick.

Custom instructions. Also in the AI tab, you can dictate the tone: respond formally, do not mention competitors, for serious faults always invite to call the support center. These apply to style, never to facts.

Face of the assistant. Choose the default icon, one of the images you find in the list, or upload your own with the Upload button. In the chat, you can also see the image of the person writing.

In what language does it respond. In the AI tab, you can check the languages in which the assistant is authorized to respond. If a customer writes in one of those, it responds in that language even if the page is in Italian; if they write in a non-selected language, the response remains in the language of the QR. Without any checks, it behaves as before: only the language of the QR.

What the assistant knows, and how much it uses it

The knowledge base: how to look inside and how to adjust it.

In the AI tab, under the Knowledge Base box, you can see how many documents and how many sections are indexed. From there, you can do three things.

See the knowledge. Opens the list of everything the assistant has on hand when responding to a customer, in the same order it reads it: documents, technical documentation, technical photos, QR photos, written instructions, features, frequently asked questions, spare parts, videos, and support centers. For each file, it states how many characters and how many sections have come out, and if anything is left out — not yet processed, processing failed, excluded by the assistant. This is where you find out why a response was vague: almost always because the right document is not in the list.

The QR tab. It is the summary that the assistant rereads before each response: what the QR is, what it is for, how it is made. If it is not there, the box says so.

Update the knowledge. Rereads written instructions, frequently asked questions, and features, rebuilds the search index, and rewrites the QR tab. Uploaded documents do not need to be reread: they process themselves when you upload them.

Knowledge power. How much material the assistant carries every time it responds. The lower it is, the faster the responses come and the more generic they remain; the higher it is, the more precise they are and the longer they take. It only applies to that QR, and from the next question.

The chat from the customer's perspective

What can someone do by scanning your QR.

From the QR page, the customer opens the assistant with the Ask the assistant button. The window can be resized by dragging the corner, and the size remains for future visits.

You can speak instead of writing. The microphone next to the field starts dictation; it stops automatically after two seconds of silence and the question is sent by itself. While the text is coming in, the message processing appears, so you can see that something is happening.

And you can listen instead of reading. In the chat header, there is a speaker: when on, all incoming responses are read aloud in the language they are written in. With the microphone, it works together — you ask by speaking and listen to the answer — and the chat becomes a phone call: it's what you need when your hands are busy on a machine. The switch remains as you left it even when reopening the page tomorrow.

• Inside each response, there is also a small speaker that reads that response aloud, just once: it's for the "I didn't understand, read it to me again" without turning anything on.
• The photos and videos attached to a response are not narrated: a voice describing an image that the listener is not looking at only makes it harder to follow. You read the response, and that's it.
• Pressing the microphone silences the voice automatically: otherwise, on a phone without headphones, the microphone would pick up the speaker.

Photos and videos in responses. If the question is about a piece for which you've uploaded a photo, the assistant shows it within the conversation: it opens in the viewer, can be enlarged and rotated. The same applies to videos you've included among the links.

Responds in the language of the question. A customer writing in German receives the answer in German, even if the document is Italian and the page is Italian.

Publish and print the QR

1. In the top right of the editor, change the status from Draft to Active.
2. Open the QR Code tab.
3. Download the PNG for printing or the SVG if the printer requests a vector format.
4. With Print stickers, you get a pre-arranged sheet.

You can change the colors of the code: the contrast between background and foreground must remain sharp, otherwise phones struggle to read it.

Advanced QR code editing. Below the colors, there is a button that opens the actual design: shape of the dots, shape of the three reference squares at the corners, color of the border and center, logo in the center, background. Next to it, there is a gallery of ready-made styles: you choose, see the preview, and save.

• Transparent background — the PNG comes out without a background, so the code can be placed on a colored sticker or a photo without the white border around it. However, the code must remain dark on light: placed on a dark background, it becomes unreadable.
• Logo in the center — a QR can lose up to about one-seventh of its surface. The site keeps the logo within that measure; if you enlarge it manually, test it with the phone before printing.

Clone a QR

For a line of QR codes that share almost everything

If you have ten models that use the same document and only differ in the code, there's no need to redo everything from scratch: clone the first one and correct what changes.

1. From My QR, click the Clone icon on the QR tab (you can also find it at the top of the editor).
2. Give a name to the copy: the suggestion is "original name (copy)".
3. Check the parts to copy and remove those that would be incorrect in the new QR.
4. Press Create the copy: the new QR opens immediately, ready to be corrected.

You can copy: QR information, photos, documents, technical documentation, technical photos, written instructions, features, FAQs, videos and links, assistant settings, options and visibility, code colors.

The copy can be in another language. In the box, choose the target language: if it differs from the original, everything that is read on the page is translated — description, written instructions, features, FAQs, attachment titles. Brand, model, codes, and units of measure remain as they are: 42.5 kg remains 42.5 kg. The text inside documents is not translated, and it doesn't need to be: the assistant already knows how to read a German document and respond in English.

References between QR and related items

In the editor, between AI and QR Code, there's the QR References tab. From there, one QR links to another of yours in two ways:

• Same tab, different language — the German version of the same article.
• Linked QR — a replacement, an accessory, the next model.

The link works both ways: linking A to B also means B is linked to A. At the bottom of the page, the published linked QR appear as Related Items, with the cover and the language flag. Only active and public ones are shown.

What it really serves for. If a person browsing in English opens a German QR and among the references there's the English version, they are prompted to switch to that. The question is asked only once per browser tab: those who respond stay here won't see it on every page.

Analytics: reading the numbers

The Analytics page shows scans, unique visitors, downloads, and questions asked to the assistant.

• Scans — each opening of the page. The same browser is recounted after 15 minutes, so simple page updates don't inflate the data.
• Unique visitors — one per device and per day.
• Downloads — how many times your files have been downloaded.
• Unanswered questions — the questions the assistant couldn't answer: they are the list of things to add.

The automatic reading. Above the graphs, there's a box where the assistant reads the numbers for you and briefly explains what's happening and what to do. It also looks at the real questions your customers have asked, so when something comes up often, it tells you. It changes by itself when you change the period or QR; with Read Again, it starts over.

Main requests. At the bottom of the page, there's the box that answers the question the numbers don't touch: not how many arrived, but what they wanted. On the left, the topics — how to use it, installation, malfunctions, maintenance, spare parts, warranty, technical data, purchase; on the right, the recurring questions, with how many times they have been asked. Those that your documents don't explain are marked to be added to the document.

How it's collected. For each question, only the topic is kept, rewritten in generic form: "the washing machine won't drain," never the customer's phrase. No IP address is saved, no reference to the conversation, no exact time — just the day. The lines cannot be reordered: there's no way to say that two questions were asked by the same person, nor to trace back who wrote.

Missed Answers

The questions the assistant couldn't answer. Each one is a gap in the document, and it gets closed once and for all.

In the dashboard and in Analytics, there's a yellow number — "6 unanswered." That number opens up, and inside are the questions as written by the customers, next to what the assistant responded. You can also get there from the Missed Answers entry in the menu, which appears only when there's at least one.

For each question, there are three things to do:

• Write the answer — it becomes a frequently asked question of the QR: customers see it on the public page and the assistant uses it to answer by itself. From the next time, that question won't come back here.
• Already fixed — for questions you've answered in another way: a document uploaded, a line added to the instructions. The thing is done and the list stops asking for it.
• Delete — not all questions deserve an answer: "how much does gasoline cost" for a boiler is not a gap in the document.

When writing the answer, the question can be rewritten: the customer's one is written in a hurry ("the thing that spins makes noise"), and as a frequently asked question, many will read it. And you choose on which QR to add it: the same question often comes up across a whole family of models, and with the checkboxes, you cover it in one go.

At the top, there are three commands: which QR, how far back to look (last 30 days or from the beginning), and See those fixed — because closed questions are set aside, not thrown away: in a month, in front of the same question, it's useful to know what was answered. From there, a question can also be put back among those to do.

Ideas

Not the numbers, but what to do with them.

If Analytics says what happened, the Ideas say what to do with it. Each idea is born from counting the facts — your customers' questions, downloads, scans — and each carries with it the numbers from which it came: "asked 12 times," "4 unanswered," "for 30 days." They are not generic advice: they are things that happened on your QR.

What kind of ideas come out:

• the same question that keeps coming back and your documents don’t answer;
• the same malfunction reported many times — when a defect keeps coming back like this, it’s usually not just bad luck for the buyer;
• a QR published without documents, which is the quickest way to lose someone who just scanned it;
• a question that would come up less often if it were among the frequently asked questions;
• spare parts that people ask for and you don’t have listed.

Every idea has what you can do (two or three concrete actions) and where it comes from (the numbers). When you’ve done it, press Done; if you don’t need it, Not needed and we won’t suggest it again.

Ideas search for themselves every now and then, in the background, and when they find new ones, a spark lights up next to the menu item. It’s not a number of things to clear out — it just says ‘something happened in here’, and it turns off as soon as you’ve looked at them. With Search for new ideas, you make them search right away.

Find your QR

In My QR, next to the search, there are two commands that are useful when the QR become many.

• Sorting — most recent, least recent, recently modified, name A-Z and Z-A, most scanned, most questions to the assistant, status, language. The sorting is in the address: the ordered list as you like can be saved among favorites.
• Cards or list — the same cards in two ways: in tiles to view the covers, in a row to scroll through many names. The choice is remembered in the browser.

Next to the name of each QR, there’s the flag of its language: with the same card in four languages, you can tell at a glance which one you are opening.

Getting found: search and your showcase

Your QR can be reached with the QR, but also by searching for them.

A sticker with a QR works when you have the QR in hand. But if someone has thrown away the box, or is looking for the document from the computer, they start from another place: Search for a QR, the public search of the site. You write the name, brand, or model, filter by category, and sort by most recent, most viewed, or name.

Every company also has its showcase: all its public QR on a single page. From the page of a QR, you can get there with the button See other QR from the seller — someone who found the washing machine document discovers that there’s also one for the oven.

For a QR to be found, two things are needed:

1. its visibility must be public (editor's Information tab): a password-protected or private QR does not appear;
2. in the Profile, Show my QR in public search must be turned on. It’s the general switch for the showcase.

You can also remove a single QR from the search while keeping the showcase on: in the Information tab, there’s the option that only concerns it. When off, the QR remains reachable with its QR and its address — it just won’t be found by searching.

Your level

The badge at the bottom of the menu: where it comes from, and what it’s for.

At the bottom left, under the menu items, there’s the card with your face, the plan, and the level. Clicking it opens the box with the score, how much is needed for the next level, and the numbers from which everything comes out. The same box is on the Profile page.

Points are not a game: they come from things that make a QR useful, and they are weighted to push in that direction.

The five levels are Newcomer, Apprentice, Advanced, Expert, and Master. In the box, there’s also the quality: the percentage of your published QR codes that are complete. It’s the most useful number of the five — twenty empty published QR codes are worth less than five well done.

• Does not deduct points. A badge that decreases would punish those who delete an old QR, that is, those who keep their things in order.
• Does not compare you to anyone. No rankings: two companies with different trades are not on the same scale.

When you level up, a notification arrives, just once.

Email and notifications

On the Profile page, choose which emails to receive:

• Email on every new scan — useful at the beginning to see if the stickers are being used.
• Weekly report — the Monday summary with scans and questions.
• Alert for unanswered AI messages — arrives when the assistant couldn't respond.
• News from ScanEron — when a new feature is released, a few lines on what it is and where to find it. Nothing more.
• Offers and promotions — deals on plans. It’s off until you turn it on.

The last two can also be turned off from the bottom of every email you receive, with the link Unsubscribe: it applies immediately and you don’t need to log into the site. Service emails — access code, address verification, support responses — will arrive anyway, because they are needed to keep the account running and are not advertisements.

To avoid filling the inbox, a maximum of one email of each type is sent every fifteen minutes.

If you don’t receive anything, press Send a test email on the same page: it will tell you immediately if the problem is with the hosting. The weekly report needs daily maintenance to be set up on the server.

Your account security

The password, the code at each access, the profile images

The Security tab of the Profile holds three things: the password, the way to log in, and the accesses that are open.

Change the password. You need the old one, then the new one twice. While you type, a bar shows how strong it is and the eye icon reveals it if you want to read it again. The field is never pre-filled, not even if the browser tries.

Code via email at each access. It’s the most valuable checkbox, and it’s usually off. By turning it on, after email and password, the site asks for a six-digit code that it sends you immediately: until you enter it, you’re not in, and knowing your password is no longer enough — they would also need your email inbox.

• The code is valid for ten minutes and only once.
• You have five attempts: after that, the code expires and a new one must be requested.
• If the email doesn’t arrive, you can request a new one from the same page.

Profile photo and company logo. They are higher up, on the same page. The Change, Edit, and Remove buttons are next to the image and go below only when the screen is narrow. Edit opens the editor: you can move the image, enlarge or shrink it, and if it doesn’t fill the square, you can choose the background color to put behind instead of leaving it empty.

Write to the staff

Assistance within the site, without opening the email

If something isn't working or you have a doubt, open a message: a person will respond, and the reply will come here as well as by email.

1. Press Messages in the left menu, or Contact the staff at the top of a QR tab (so we know right away which QR you are talking about).
2. Write a clear subject and calmly explain the problem.
3. If needed, attach screenshots, PDFs, or documents.
4. Press Send to staff.

When we reply, a small red number appears on the bell at the top right: from there, open the conversation and you can continue writing. Each conversation stays open as long as needed, and if you write again inside a closed one, it reopens automatically.

Spare parts and service centers

The code to dictate over the phone, and who to call

The spare parts. In the Contents tab, next to Documentation, Technical Documentation, and Technical Photos, there is the Spare Parts section. Each spare part has a code, a brief description, and up to three photographs.

There are no prices or availability, and that's intentional: anyone scanning the QR with the broken part in hand wants to know one thing only, the code to dictate when ordering. The photographs help ensure the right part is taken.

1. Open the QR and go to Contents.
2. Press the Spare Parts section: the file box gives way to the list of spare parts.
3. Press New spare part, write the code as it appears in the catalog and a brief description.
4. Save: the box stays open and the photo area lights up. Upload up to three.

The service centers. In the Service Centers tab, you indicate where to order spare parts and who to call for a repair. On the public page, the contact details become buttons: from a phone, you press "call" and the call starts.

The registry of the centers is yours and applies to all your QR codes: you write a center once and then add it where needed by searching for it by name or city (two letters are enough). The day it changes phone number, you correct it once and it changes on all tabs.

Rights on the material you upload

Who is responsible for what, and what happens if someone reports

A document is a protected work like a book or a photograph. When you upload it to ScanEron, you are publishing it online, and the responsibility for having the right is yours: we provide the infrastructure, we do not check in advance what comes in.

For this reason, the first time you save or have the assistant read newly uploaded files, a box appears with the list of files and a checkbox:

The confirmation is recorded with date, time, and the files it refers to. It’s not bureaucracy: it’s what, if someone contests a document in the future, indicates who uploaded it and when.

What you can upload without thinking. The documents of the QR codes you sell or assemble; the photos you have taken; the texts you have written; the documentation that your supplier has authorized you to distribute — and if you’re not sure that the authorization exists, an email to the supplier takes ten minutes.

What is better not to upload. The document of another brand taken from the internet, even if the QR is identical to yours; photographs taken from someone else's catalog; technical sheets of a competitor.

If someone reports your QR. At the bottom of each public page, there is a link to report a violation. If a report comes in about your QR:

• the documents of that QR immediately stop being downloadable, pending verification;
• the page remains online: photos, features, and assistant continue to work;
• you receive an email telling you what happened;
• we verify within 24 hours. If the report is unfounded, we restore everything as it was, without you having to do anything.

The full rules are in Article 7 of the Terms of Service.

Languages and appearance

The site is available in Italian and English: you can switch from the flag at the top right on every page, even without logging in. The choice is saved.

Each QR has its own language, chosen when created from those active on the site. It’s the language in which what you put inside is written — the document, the FAQs, the names of attachments — and it’s shown with the flag in My QR. It’s not a cage for the assistant, which continues to respond in the language of the person writing to it.

The light or dark theme follows your computer or phone settings; you can force it with the switch next to the flag.

Plans and limits

When a limit is reached, the site clearly states it and the already published content remains online: nothing is lost.

Frequent issues

The QR is not readable. Print it larger (at least 2.5 cm on each side), increase the contrast between the two colors, and avoid very curved or reflective surfaces.

The assistant says it can't find the information. Check in the AI tab that the document is Processed. If it’s still pending, press "Save and process" again. If the PDF is a scan, reading will take longer.

A document is not being read. If it’s a very old file (.doc or .xls from the 90s), try reopening it and saving it as DOCX, XLSX, or PDF: reading will be much more accurate.

I'm not receiving emails. Check your spam folder, then use "Send a test email" from the Profile.

I've exceeded the available space. Delete files you no longer need or upgrade to a higher plan. Already processed documents remain indexed for the assistant even after freeing up space.

I changed the document: do I need to reprint the sticker? No. The QR address never changes, so you can replace the content as many times as you want.