Structurization, Retrival Augmented Generation and AI Tools with Lang Chain

LangChain is ecosystem that introduced unified interface for multimodal conversational models. It provides also ecosystem for models efficiency of llsm deployed on production apps.

We will learn basics of LangChain by 3 examples.
- response structuring - on input we will put pdf invoice, and data model. We expect that out program will be able to get different PDF invoices and extract from them client, date, invoice number and total value for payment. This can be useful in digitization of data that are received in form like images of PDF.
- rag - in input we will put book and detailed question about details of one chapter, we will compare results from RAG that we created and Chat GPT. This can be used with any type of text, like documentation, or offer of our company.
- tools - as input we will give database and questions about data in human language. We expect from model to introspect db structure, build queries dynamically execute them and present summary of results.

https://python.langchain.com/docs/integrations/tools/sql_database/

What do you need:
- basic knowledge of python
- laptop

What you will learn:
- how to use langchain
- what are these capabilities

You don't have to be experienced python user, enough to know how to install packages and run python hello world. There is assumption that it is your first contact with langchain.

Event is free, but to use these tools you will probably have to spend about 5 USD for your own API KEY from any LLM provider. During even we can provide you our keys.

To prepare you can read more about apis that will be presented:
https://python.langchain.com/docs/how_to/document_loader_pdf/#other-pdf-loaders
https://python.langchain.com/api_reference/upstage/index.html#langchain-upstage-tools

📅 Date & Time: Thu, Apr 20, 2025 01:00 PM (2 hours)
📍 Location: LendSpace · 4 Sherif Khimshiashvili St, Batumi 6000, Georgien · Batumi