Michał Kurkowski


Constructing AI-Driven Systems Utilizing Large Language Model (LLM) Frameworks

#Introduction - about 45 seconds

Reference to popular culture, e.g. "2023 is like a science-fiction scenario. Machines learned to talk and AI technology became commonplace."
Describing the benefits and opportunities resulting from the use of LLM, as well as potential threats.
The introduction thinks that this technology is not just huts, but is increasingly an integral part of companies.
Highlighting the fact that acceleration is on the horizon and companies that don't see it may be falling behind.

#Business risk - about 30 seconds

Emphasizing the need to consciously use AI technology, paying attention to the risks associated with dependence on suppliers such as OpenAI.
Proposing a smart strategy of investing in your own self-hosted AI that you can control and customize to your needs.

#How to create LLM-powered tools - 2 minutes 30 seconds

A quick introduction to what LLM is as a "black box", how to communicate with it via API and how to use it to create AI tools.
Presentation of LangChain - description of basic concepts in an interactive way, showing what the process of creating and implementing tools looks like.
An example of creating an AI agent is presented in the form of a story. You can describe a scenario where an agent is tasked with responding to RFPs and show how he uses various tools to achieve his goal.
Highlighting that what you just described is the foundation of a new area known as FMops that is worth further exploration.

Completion - about 30 seconds


explaining that the presented tools and technologies are not the future, but the present.
Encouragement to explore LLM opportunities and actively participate in the AI revolution.

Share this with a colleague