Langchain RAG Course: From Basics to Production-Ready RAG Chatbot

Build a production-ready Retrieval-augmented generation (RAG) chatbot that can answer questions based on your own documents using Langchain. This comprehensive tutorial guides you through creating a multi-user chatbot with FastAPI backend and Streamlit frontend, covering both theory and hands-on implementation.
Topics covered:
RAG fundamentals and applications
Langchain basics and LCEL
Document processing and vector databases
Building conversational RAG chains
Multi-user chatbot development
FastAPI integration
Streamlit interface creation
By the end of this video, you'll have built a functional AI chatbot capable of answering questions based on custom data sources.
For code samples and additional resources, visit our blog: blog.futuresma...
#Langchain #RAG #ragchatbot #llm

[@Pratik345-b1y]

This is the best explanation of RAG ever given by anyone - Detailed + Beginner to Advanced. One suggestion, we need a playlist for Gen AI and Agentic AI to follow in sequence, currently it's hard to navigate over youtube channel !!

[@FutureSmartAI]

Great suggestion!

[@santudam35]

One of the best video i came across to learn the LangChain and RAG to build the AI Application from basic. Thanks for the the awesome work !!

[@FutureSmartAI]

Glad it was helpful!

[@GuaravBansal-l2y]

This is best production level RAG code. That is what I was looking for.

[@Pratik345-b1y]

Please keep uploading more videos/tutorial like this to share your knowledge - Its very much helpful or if possible please launch a course on Gen AI + AI Agent as well !! You are doing an fantastic job - Reaching million subscribers soon 💯

[@Mostafa_Sharaf_4_9]

I really missed your videos, and now you have come back with a comprehensive and wonderful video that solved a lot of my problems, so thank you and I hope you continue.

[@FutureSmartAI]

Great to hear!

[@GaneshMandpe-y6w]

This is the best video on KZbin if anyone wants to learn RAG. Great job Pradip, very very impressive work mate!!

[@FutureSmartAI]

Thanks for the kind words! I put a lot of effort into making this video informative and helpful.

[@aidev8926]

Please keep uploading like this exquisite content. Thank you!

[@AshisRaj]

Excellent work brother. Very lucid explanation. It will even help the season developer to refine their understanding of the concepts. Looking forward for more such videos like these. You have earned my respect and a subscriber :).

[@FutureSmartAI]

Glad it was helpful!

[@pravinthombal9787]

excellent work

[@FutureSmartAI]

Thanks for the visit

[@rajudk9132]

Timestamp RAG: 31:27 Embedding: 38:07

[@vivekshindeVivekShinde]

Much needed. What an explanation. Thanks a lot 🙌🏻

[@FutureSmartAI]

Glad it was helpful!

[@sadiasiddique6254]

Great video! Could you please create one on combining fine-tuning with Retrieval-Augmented Generation (RAG) for chatbots? Fine-tuning can be costly for certain use cases, but applying it selectively to establish the tone or behavior of RAG models could be highly efficient. This would be useful for instances where we want the model to follow a specific conversational style without extensive, full-scale fine-tuning

[@alexramos587]

Nice series. Please keep uploading.

[@satyamchoudhary2573]

bhai, what an explaination! just wow

[@sangram7153]

Excellent

[@duetplay4551]

Nothing...Just log in to say "Merry Christmas and Happy New Year!" to Professor Pradip and other "classmates"🥳

[@FutureSmartAI]

Merry Christmas and Happy New Year to you too!

[@yazanrisheh5127]

Hello Pradip, could you make a video on how can we split based on the ranking of the employee in an organization? For example executives would have access to financials but a junior would have access to a guide documentation

[@rajank2909]

very important and great videos please keep posting , how can we store unstructured data like table and image in vector db

[@honor1498]

Awesome sir thanks a lot I have a request sir please make a vedio on integrating diagram generation feature also in this chatbot sir it helps alot sir

[@Net4XInnovation]

This is the best course about RAG and langchain i saw, thank you so much. how about if we need to connect a relational DB as a RAG source instead of documents uploaded, do you have anything about this please?

[@FutureSmartAI]

Thanks check this tuorial kzbin.info/www/bejne/nKTWZ3aooraIaLssi=VHW6PUqhgtofJw_8 If you want to combine RAG and NL2SQL in single agent check this blog.futuresmart.ai/multi-agent-system-with-langgraph

[@eige2992]

Hello, I've encountered an error when I try to embed the documents. Error Code: 403 Project doesn't have access to model 'text-embedding-ada-002". Can you help me solve this?

[@Danyal_alam]

it was such a great course. We can deploy this application on azure or aws also ? not only streamlit right?

[@FutureSmartAI]

Yes you can deploy this app anywhere

[@iamsatyajit963]

Best explanation and the code walkthrough is amazing. Bdw, where can i get the documentsthat you have used in the code?

[@FutureSmartAI]

Link in the description that has code and docs

[@saikrishna235]

Great explanation! Where can I find the code/notebook present in the video?

[@FutureSmartAI]

link In the description

[@satyamchoudhary2573]

please make video on all topics in same way

[@robrita]

thank you!!

[@Pratik345-b1y]

Getting this error : openai.OpenAIError: The api_key client option must be set either by passing api_key to the client or by setting the OPENAI_API_KEY environment variable : Any idea ?

[@celestialgamer360]

if you don't have openai api key you will often counter error so, try to use open source model

[@mithunmahato309]

One thing i can't understand is : what is context ? In the example 52:00 you put all the document as context.

[@FutureSmartAI]

The context consists of relevant documents retrieved from a vector database. These documents are processed to extract only the page content, excluding metadata, and all content is combined into a single long string, with each document separated by two newline characters.

[@tajudeensanni9240]

Nice one! please, the blog post seems not to be opening.

[@xxtinct_seldom_species]

How can be save tokens. every time we hit LLM or open AI. we consume some tokens. and token are pricey right? how do we save them ?

[@rajatsrivastava9353]

my langsmith account is showing: Failed to execute 'getReader' on 'ReadableStream': ReadableStreamDefaultReader constructor can only accept readable streams that are not yet locked to a reader || how to resolve this?

[@ArunKumar-bp5lo]

Finally

[@umer_c0des330]

In newer version of langchain v0.3, these chains has been deprecated, in favor of the more flexible and powerful frameworks of LCEL and LangGraph. Then, why do you prefer to use these chain??

[@FutureSmartAI]

I've also demonstrated in the video how to achieve similar functionality without this specific chain, using LCEL instead. Additionally, I don’t believe these chains have been fully deprecated; even the 0.3 documentation still includes them. You can check it out here: python.langchain.com/docs/tutorials/qa_chat_history/

[@Pratik345-b1y]

Hi @pradip - How can we integrate voice agent functionality in this ? like if customer calls, our ai agent should speak to them as real human

[@Pratik345-b1y]

Can you help pls

[@FutureSmartAI]

For voice call you will need to see provider but if you want that your agent or bot comminutae with audio and understand speech then you can do it check kzbin.info/www/bejne/jWSpkpeeftWGo7csi=-OH0eThgP8gdaUJy

[@Rakesh_Seerla]

open ai api key limit exceed when ever i have used that it shows the same 😮‍💨

[@kk008]

how to evaluate to built RAG application?

[@sanketkurle2251]

❤❤

[@prashlovessamosa]

Thanks

[@SKini18]

Can we load xlsx or xls files and chat with the file using this ?

[@FutureSmartAI]

If its xlsx you should explore NL2SQL way it will work better that treating it as text and use RAG

[@SaiLikhithaP24MCS004]

this is one amazing video sir ! I have a question, weaviate seems to give a tough competition to chromaDB so how to choose between vector DBs

[@FutureSmartAI]

Let me see if I could create video on that

[@Magmatic91]

Thank you for this great tuto, I have question though : The app you developed is on localhost, but how we can deploy so that it's available online ?

[@FutureSmartAI]

Its like deploying any api check this kzbin.info/www/bejne/bXe5gaFoatKge7Msi=k9IiN8XS13hn_O48 kzbin.info/www/bejne/b2GXlIpvoa9qgrMsi=aydaAzyj-nyJPDke

[@Magmatic91]

@FutureSmartAI thank you

[@pravinthombal6742]

If we provide chat history every times, with query then token size will also increase, so is it okay? Please answer

[@FutureSmartAI]

Yes we need to pass chat history since llm dont remember anything but there are ways to limit chat histry check this python.langchain.com/docs/how_to/trim_messages/

[@abhishekpawar7170]

27:36

[@chaithanyavamshi2898]

Please teach Huggingaface or ollama open source models instead of Open AI LLMs

[@FutureSmartAI]

Planning open source models in next video. I work with multiple clients and they still prefer open ai models rather spending money hosting open source models

[@chaithanyavamshi2898]

@@FutureSmartAI Thank you for the reply. Interesting to know and company which I work for want to host open source models due to privacy and security and no need to worry about vendor lock in. Can you please guide or have any plans using litellm to make the code model agnostic?

[@lithinullal8756]

great video! can you share the code used?

[@FutureSmartAI]

In the description

[@chimwemwechinamale6716]

Hi Pradip. Is it still worth it to pursue this space/market on upwork? is there still demand for it ?

[@FutureSmartAI]

Yes

[@abhishekpawar7170]

1:15:05

[@mithunmahato309]

Can i use gemini pro instead of open ai ?

[@FutureSmartAI]

Yes

[@ArpitYadav-ws5xe]

wonderful...you are doing excellent.. can we get this code file?

[@FutureSmartAI]

Yes. code and explanation link in description

[@Pratik345-b1y]

@@FutureSmartAI code isn't complete - can you publish it on git or add google collab link

[@FutureSmartAI]

@@Pratik345-b1y All code is availble on hashnode series whose link is in description. Blog will also have link to original notebook

[@Pratik345-b1y]

@@FutureSmartAI Got it - Thank you so much - Your work is really awesome !! Please create more content - its more valuable than any paid courses

[@imrozalam9042]

@@FutureSmartAI i am not able to find

[@RajatKUMARNayak-ce6ki]

you should sahre the github Repo as well

[@FutureSmartAI]

yes check link in description

[@awadheshamar6012]

Thanks a lot for such a wonderful content. Could u please list down all step required to get code in local and make it running 1. Clone code using git clone 2. Replace api key 3. Run command for Backend 4 run command for fronted This will help to play around with this code

[@awadheshamar6012]

How to run both application

[@awadheshamar6012]

I m new to python

[@shobhitagnihotri416]

I think It would be better to revise

[@FutureSmartAI]

Thanks for suggestions

[@anonymouscaveman8557]

lottery lag gai

[@nothing-t4c1g]

Hi @pradip

[@bhaveshxrawat]

19:47