The development of large language models (LLM) has made this possible. There are many new models around, such as ChatGPT, Gemini, Bard and others.

Basically, it’s possible to ask a question in natural language and you get an answer in natural language. As simple as that.

It works for all industries and topics. Such as in Healthcare, where countless guidelines need to be observed, or in Mobility, where instructions depending on the case need to be checked, or in Finance where regulation and compliance require a lot of instructions need to be observed.

What obstacles could I encounter and how can I work around them?Like many things in life, this is also not that easy. The publicly available models use publicly available information and data that has been fed into them, either on purpose or by not being aware of it. And that is exactly where some of the hurdles for easy use lies.

In a business context information searched for is often confidential, for internal purpose only or even protected by a patent. To name a view of the reasons not to use a public model for most of the business needs.

There are large language models available as open source, already pre-trained with carefully curated data from one or multiple specific domains. To make it usable for your own purposes and most important keep your information you feed in private, the model needs to be fitted with your own input

Wasn’t there something about privacy earlier in this text? Yes. And this is why you should not put your information in a publicly available model, but in one that is exactly tailored for your needs. The good news is that Netcetera has already prepared the ingredients for this.

We have created a tool called DocDive, based on Generative AI, which allows exactly to handle your privacy but benefits from open-source models.