LangchAIn

What do you like best about Langchain?

Langchain is the best framework for building RAG applications its supports all find of Large Language Models that is both open-source like llama, mistral and closed-source models like OpenAI, and Anthropic using their Access token. It also supports using local LLMs using the Ollama. I build RAG application for our enterprise data using it , its very simple to build and it has many features like a chain of thoughts and memory. They have clean documentation which we can refer to and add more features. We can easily integrate with their other products like Langsmith and Langgraph

What do you dislike about Langchain?

I didn't feel any downside while using the Langchain but one thing is they have many version depecdences which will through error if you don't install the correct version.

What problems is Langchain solving and how is that benefiting you?

I was asked to build an RAG Module on our enterprise data so that our company employees can make use of it to search and get information from our database like documents,PPTs,pdfs etc.. Using Langchain I was able to build it using the opensource LLM model like Llama 3.1 8B, their documentation made it very easy to refer and build. And using the Langsmith which is their other product which helped me for productionization of our enterpirse RAG.I felt Langchain is very easy to implement compared to others like LlamaIndex.

What do you like best about Langchain?

i really had fun building RAG with Langchain, the option it provides are really amazing it supports mullitple vendors model for llm for example openai, oolama, mistral ai if you want to go with open source models ofcourse huggingface is there for us well langchain support that as well, the implementation is really easy and about the documentation it is really good straight at point even a basic python language understanding coder can start with langchain in no time, i had integrated langchain with langflow that is also an amazing open source product

What do you dislike about Langchain?

Well i have no particular dislikes for langchain but as a beginer in langchain i had issue with respective dependency conflict between langchain and langchain community library and other dependecy conflicts other than that i think i have not faced that many issues such that i can say as my disliked towards langchain overall really amazing work from langhcain community

What problems is Langchain solving and how is that benefiting you?

So recently i created the product using langchain, so the issue was the employes have to do hard work to find a specific id through various types of documents so i had created a RAG pipeline where instead of doing all the hard work and just ask the question what they required from the documents with all metadata like pagenumbers etc.., this solved the problem big thanks to langchain comunity for this opensource

What do you like best about Langchain?

It is one of the best packages required to use Large Language Models in the field of Generative AI, it is easy to adapt and works like a charm and it keeps upgrading itself to be compatible with the latest technology

What do you dislike about Langchain?

Sometimes some feature might not be present in the latest version of langchain which was previously there, so we have to rewrite our code to match with the new version, if they could just support the older versions as well then it would have been better.

What problems is Langchain solving and how is that benefiting you?

It is the primary package I use to develop Generative AI based applications, it does everything related to that field so it solves most of the problems faced in the field of LLMs and Generative AI.

What do you like best about Langchain?

now I can Connect Any number of LLMs with any Number of Tools. I am making Agents using Multiple Prompts. I can create and add memories to my conversational chain. I can read from PDFs as well as databases with different vector databases. I can also integrate LLMs like OpenAI, Mistral, Llama, etc with the Internet as well as with APIs It gets Additional Data. It's Easy to Use and Implement in code.

What do you dislike about Langchain?

I feel like the code written in Python for Langchain makes it a little slower. Also, there are restrictions in using OpenAI or Claude Function calling with langchain. Also, there are better faster solutions like Haystack.

What problems is Langchain solving and how is that benefiting you?

Langchain integrates APIs, Tools, InMemoryCache, and supports multiple Agents, Multiple LLMS, Multiple VectorDBs, Multiple conversations, and retrieval chains. Langchain helps to make AI/ LLM Agents that can Work together in research, in automation as well and in creating solutions using tools that can connect to the internet as well as to databases. We can make our own custom LLM that can work on provided data using open or closed-source LLM's APIs.

What do you like best about Langchain?

Easy Handel and productive options for usage made faster execution

What do you dislike about Langchain?

Nothing as of now as we use it didn't face any issues

What problems is Langchain solving and how is that benefiting you?

Legacy way of handling LLM is handled now with langchain