There is a lot of talk about artificial intelligence, but for a small or medium-sized business the question is practical: in which everyday activities can it really save time? The answer lies not in generic promises, but in specific use cases tied to problems every company knows.
We have gathered 10 use cases based on the features of ThorIA, Thorsoft's business AI. For each one you will find the problem, how AI solves it and what you need to make it work.
1. Analysing PDF, Excel and image documents
The problem. Contracts, invoices, certificates and tables arrive in different formats, and reading them one by one takes time, especially when you need to extract a few pieces of information from long documents.
How AI solves it. The AI reads PDFs, Excel files and images, summarises their content and answers specific questions, for example what the main terms of a contract are or which items exceed a certain threshold in a spreadsheet.
What you need. The documents to analyse and clear instructions on what to look for. Answers must be checked by a person when they are used for important decisions.
2. Checking master data in documents
The problem. A document uploaded by a supplier or employee may refer to the wrong person or company: an error that often only comes to light during an inspection.
How AI solves it. The AI compares the details in the document, such as the company or person name, with those of the party the document is linked to, and flags any inconsistencies.
What you need. Up-to-date supplier and employee records and an archive in which every document is linked to its party.
3. Calculating expiry dates automatically
The problem. Many documents have limited validity, but the expiry date is not always stated explicitly. Calculating it by hand for each document is slow and error-prone.
How AI solves it. The AI determines the expiry date based on the type of document and proposes it to the operator, who validates it before it is recorded.
What you need. Well-defined document types and an operator responsible for validation.
4. Sending automatic email reminders
The problem. Chasing suppliers and employees for missing or expiring documents is repetitive work that grows with the number of parties to manage.
How AI solves it. The AI sends email reminders to suppliers and employees when a document is missing or about to expire, with wording suited to the situation.
What you need. Correct email addresses, rules on when and how often to send reminders, and an up-to-date register of expiry dates.
5. Generating communications, reports and quotes
The problem. Writing emails, periodic reports or draft quotes takes time that could be spent on higher-value activities.
How AI solves it. The AI prepares a first version of communications, reports and quotes based on the company's data and templates, which a person then reviews and completes.
What you need. Reference templates, a defined tone of voice and human review before sending.
6. Querying data in natural language
The problem. Getting information out of business systems often means knowing filters and reports, or asking an expert colleague.
How AI solves it. You ask questions in plain language, such as "which suppliers have documents expiring this month?", and the AI answers based on the data available.
What you need. Structured, accessible data and permissions that set what each user can see.
7. Applying policies and content controls
The problem. Without rules, AI can be used for unintended purposes or produce content inconsistent with company policies.
How AI solves it. Company policies and filters define what the AI can do, which sources it uses and how it behaves, applying content controls.
What you need. An internal AI usage policy, translated into rules configured in the system.
8. Setting up scheduled automations
The problem. Some activities repeat at regular intervals: weekly checks, monthly summaries, periodic reviews.
How AI solves it. Scheduled automations carry out these activities on a schedule, without anyone having to remember them.
What you need. Well-defined activities, a set frequency and someone responsible for checking the results.
9. Creating a shared knowledge base
The problem. Procedures, instructions and reference documents are scattered, and colleagues keep asking the same questions.
How AI solves it. A knowledge base with shared files lets the AI answer questions based on the company's documents.
What you need. Up-to-date reference documents and someone to maintain them.
10. Managing roles and permissions
The problem. Not everyone in the company should see or change the same information, even when using AI.
How AI solves it. Roles and permissions, for example viewer, editor and owner, determine who can consult, edit or administer content and settings.
What you need. A map of company roles and clear rules on who can access what.
How to get started
You do not need to activate all use cases at once. The most effective approach is to choose one or two problems that weigh most on the organisation, for example supplier document expiry dates, and start there. Once you have the first results, it is easier to extend use to other activities.
Some conditions apply to every case:
- Well-organised data: AI works best on well-organised documents and records.
- Clear rules: policies and permissions set what the AI can do and for whom.
- Human review: important results must always be checked by a person.
- Data protection: knowing where the data is and whether it is used for training.
Many of these cases, such as master data checks, expiry date calculation and reminders, are already built into the IntraThor DMS document management system. To find out more, also read AI and document management: automated checks and our checklist for choosing business AI.
With ThorIA, data is hosted in the EU (Milan, Italy) and is never used to train models; users are unlimited and the Thorsoft team tailors the system to business processes. Thorsoft classifies ThorIA in the limited-risk category of the AI Act.
F.A.Q. - Frequently asked questions
What is business AI?
It is an artificial intelligence system configured on a company's data, rules and processes, to support activities such as document analysis, expiry dates and communications.
Which use case should I start with?
With one or two problems that weigh most on the organisation, such as managing supplier document expiry dates, and then extend its use.
Does AI replace people?
No. AI reduces repetitive work, but important results must be checked by a person who remains responsible for decisions.
What does an SME need to use AI?
Well-organised data, clear rules on what the AI can do, defined roles and permissions, and human review of the results.
Does ThorIA use company data for training?
No. With ThorIA, data is hosted in the EU (AWS Milan, Italy) and is never used to train models.
Request a demo
Find out which ThorIA use cases best suit your company.
General information only, not legal advice: always check the current legislation.