Enterprise AI & Strategy · Use cases
Generative AI: 6 Use Cases for Companies
Generative AI writes, summarises, reads documents and answers questions. Six use cases show which applications actually take work off people’s hands in companies, most of them with a project already in use.
Julia Rose
Key takeaways
- Generative AI creates new content such as text, images or code based on large language and image models.
- The six most common uses: summarisation, information extraction, assistants, code, text classification and content creation.
- Project examples: 80% less manual effort in data extraction, call notes written automatically in an MVP.
- A demo is quick to build. Production needs data connection, integration, quality checks and operation.
What generative AI is
Generative AI creates new content based on what it has learned. Large language models write and analyse text, diffusion models create images. For companies, language matters most: models understand requests in plain words, read documents, pull out information and write answers in any tone and format. The technology delivers its full value once it is connected to your own data and systems.
6 use cases for generative AI
1. Summarising texts and conversations
Long reports, market analyses, contracts or call transcripts become short summaries in a fixed format. Experts grasp the essentials without reading every document in full.
In practice: arvato
In the call centre, a note with intent, category, actions and next steps is created automatically after every customer call. In an MVP, call notes are written automatically.
More in our article Generative AI for summaries.
2. Extracting information from documents and emails
Dates, amounts, names, shipment or order details sit in emails, PDFs and contracts, often in changing formats. Language models read them and pass them in structured form to ERP, CRM or specialist systems.
In practice: Fr. Meyer’s Sohn
The logistics company receives thousands of emails a day in German and English with no fixed format. A GPT-based pipeline on its own servers extracts the shipment data and cuts manual effort by 80%.
More on this in the article Email inbox automation.
3. Assistants for everyday work
AI assistants, often called copilots, search documents, summarise, draft emails and answer questions from company data. They exist as standard products such as Microsoft 365 Copilot or as custom solutions that access systems like SAP.
In practice: car manufacturer and GS1 Poland
For a premium car manufacturer (under NDA), an assistant answers questions about planning data from SAP BW by text or voice, in seconds instead of hours. At GS1 Poland, we introduced employees from several departments to Microsoft 365 Copilot for their own tasks in a workshop.
Read the SAP planning assistant case study · Read the GS1 Poland case study
4. Writing and reviewing code
In software development, AI tools suggest code, explain existing code, write tests and find bugs. Well-known examples are GitHub Copilot, Tabnine and Amazon Q Developer, which absorbed the former Amazon CodeWhisperer. Developers remain responsible for the architecture, review and security of the code.
5. Classifying texts and detecting sentiment
Language models and NLP methods assign texts to categories, recognise technical terms and assess sentiment, for example in customer feedback, tickets or social media.
In practice: pharma and events
For a global Swiss pharmaceutical company (under NDA), an NLP pipeline finds adverse drug events and sentiment in medical texts and social media, in five languages. For the conference organiser Re-Work, a real-time tool scores sentiment on social media while the event is running.
Read the pharmacovigilance case study · Read the Re-Work case study
6. Drafting content
Generative AI helps draft texts, product descriptions, presentations and images: it delivers outlines, first versions and variants in different tones. People check facts, brand voice and image rights before anything is published.
From demo to production
A first demo with an available model can be built in a few days. A system that works reliably every day needs more:
- Data connection: access to the right documents and systems, with the right permissions.
- Integration: results land where people work, such as ERP, CRM or email.
- Quality checks: test sets with real cases and ongoing evaluation in production.
- Operation and data protection: cloud, own servers or hybrid, depending on the data requirements.
A process analysis shows which use case pays off first. What a structured start looks like is covered in our article How companies implement AI successfully.
Which use case fits you?
We look at your workflows and assess where generative AI takes over the most work and what a first step costs.
Request a process analysisFrequently asked questions
AI that creates new content such as text, images or code. For companies, large language models matter most: they read documents, extract information and write answers.
Common ones are summarising texts and conversations, extracting information from documents and emails, assistants for everyday work, coding support, text classification and sentiment analysis, and drafting content.
Considerably more than the demo. Data connection, integration with existing systems, quality checks and operation make up most of the work.
At theBlue.ai the analysis is available at a fixed price from €3k. Implementation typically runs between €25k and €100k, with a scoped estimate at the end of the analysis.
Julia has been part of theBlue.ai since 2019 and has accompanied the development of AI applications in the enterprise environment since the company’s early days. In her role as Marketing Lead, she works closely with the engineering and consulting teams and makes complex technical topics understandable and accessible for decision-makers.
In her articles, she writes about practical experience from enterprise AI projects, as well as the challenges and opportunities of using AI in companies.