Analyse documents
PDFs, data sheets, invoices and emails are read and turned into structured data.
Maximum intelligence. Fully automated. AI systems by core.iq
Maximum intelligence. Fully automated. AI systems by core.iq
Our automations embed AI models directly in the process chain. They read documents, check content, decide according to your rules and complete the next step themselves.
AI automation pays off as soon as the same documents, texts or enquiries are handled by hand every day. Typically businesses with 2 to 50 people, where a single person carries most of that work.
We do not start with a large project, but with one single process. It runs within days, and you decide from the result whether the next one is worth it.
This is probably not for you if the task only comes up a few times a year. Doing it by hand is cheaper then.
Invoices and delivery notes are typed in by hand.
AI reads them, matches them against the order and files them ready to book.
Every enquiry in the inbox is read and answered one by one.
Enquiries arrive sorted, with a draft reply for you to approve.
Product texts and translations get written in the evening.
New items reach the shop with text, attributes and translation.
Wherever content needs to be understood, assessed or created, AI plays to its strengths.
PDFs, data sheets, invoices and emails are read and turned into structured data.
Product texts, FAQs, translations and social media posts are written in your tone of voice.
Assign categories, verify mandatory information and detect deviations, even in messy data.
Adjust prices, grant approvals or ask a targeted question when a case is unclear.
Customer enquiries are classified, answered or forwarded to the right person.
Cut out and standardise product images and generate new visuals from product data.
Four figures from field studies and official statistics, each with its source. None of them are our own estimates.
35 per cent have built it into their processes, 39 per cent are experimenting. The gap to those still watching grows every year.
AXA SME labour market study 2026, published on 15 September 2026For staff with little experience the figure was 34 per cent. Answer quality rose measurably as well.
Brynjolfsson, Li and Raymond, “Generative AI at Work”, NBER Working Paper 31161, 5179 staffPlus 12 per cent more tasks completed and quality rated more than 40 per cent higher than in the control group.
Dell’Acqua and others, Harvard Business School Working Paper 24-013, field experiment with 758 consultantsOne hour of manual work on every working day is around 14,000 francs a year. An automation is usually paid off within the first year.
Swiss Federal Statistical Office, labour costs per hour worked, published via kmu.admin.chThe flip side: in the same Harvard study, participants were 19 percentage points more likely to get it wrong on a task that sat outside what the model reliably does. That is exactly why we build every automation with checks and firm limits instead of blind trust. How the guard rails work →
Rules work until an exception comes along. At that point, an AI model takes over in our systems and completes the process.
Works until an exception comes along.
AI takes over where rules fall short.
The AI works like an experienced team member: it reads the documents, notices what is missing and completes the rest according to your specifications. Every step is logged and traceable.
Two-page data sheet with nutrition table analysed. Ingredients, nutritional values and allergens are captured in a structured way.source: supplier_datasheet.pdf
Belongs to sports nutrition. Milk must be declared as an allergen. French and Italian titles are still missing.rules: shop_guidelines · food_law
Create item, set allergen notice, generate translations. The retail price is within the margin target, no approval needed.margin: within target
The item is live in the shop in three languages and the stock is recorded in the warehouse. The social media post about the new product is scheduled.duration: 47 seconds
Our AI doesn’t act freely but within fixed rules that we define together with you. Every decision is logged and remains traceable.
Where an approval makes sense, the system asks instead of guessing. This keeps quality high, even at large volumes.
We are not tied to any provider. For every task we choose the model based on quality, speed and cost. When a better model comes onto the market, we switch without anything changing for you.
Claude
OpenAI
DeepSeek
Geminiand other providersTexts, image analysis, translations and decisions can be handled by different models depending on the requirements.
Items, attributes, nutritional values and translations are created automatically from data sheets and manufacturer websites.
Customer enquiries are read, requirements captured and a quote prepared that only needs to be approved.
Invoices and expenses are read, assigned to the right account and transferred to accounting.
Ingredients, limits and labelling are compared with the applicable regulations and documented.
Images, captions and weekly plans are created from the product range and published after approval.
Common questions are answered directly; everything else goes to the right person with a summary.
Where this comes from, what it costs and who owns the data.
We put AI where it has to work every day: in receipts, product data, texts, translations and the accounts. The same models and the same checks run in systems we built for clients and that are in daily use there. In online retail we cover the whole path from product data through to the booking.
A fixed price per process, instead of billing by the hour. On top of that the model usage, billed by volume. Support optional and cancellable monthly. The binding quote follows after a short look at your process.
Before the first run we define which data reaches a model at all and which stays in house. We use business accounts where your content is not used to train the provider’s models, and on request everything runs in your own account. As a processor we work under a contract based on Swiss data protection law, and every action is logged.
The most important answers at a glance.
Do you have a task in mind that involves a lot of manual work today? We’ll show you how an AI model can take it over and where checkpoints make sense.