ADVANCED Python AI Multi-Agent Tutorial (RAG, Streamlit, Langflow & More!)

[@TechWithTim]

Get started with DataStax: dtsx.io/techwithtim-astradb Langflow Github Rep: dtsx.io/techwithtim-langflow Check out DataStax AI PaaS: dtsx.io/techwithtim-aipaas

[@nobleman-arts]

I am a computer engineering student, and you have been making me to stand out, thanks!!

[@TechWithTim]

That’s awesome!

[@KumR]

Before I see the video, can I ask if your agent checks weather?

[@TechWithTim]

@@KumR they do not

[@shamirshakir]

Finished the complete tutorial. Can't thanks enough for this wonderful tut.

[@amy4smiles]

Thank you! I'm over half way through and you explain it all so well!

[@SamBeera]

Great video, thanks for creating this and step thru to help many build agentic workflows.

[@MardaPoudel]

aitutorialmaker AI fixes this. Multi-agent AI tutorial overview.

[@douxdoux84]

Hi Tim, love all your videos. Thanks for the great content and outstanding teachings❤❤❤

[@karjedav]

Great tutorial! Quick question: how do I put this tool on the internet so that other people can use it?

[@kaustubhgupta12]

Hey Tim, great video. I wanted some suggestions from you. As you have already build the fitness tracking app, any ideas or suggestions for what else I can build on the same lines of using multiagent LLMs?

[@viveksuryavanshi6165]

wow, been waiting for this one.. Thanks much 🙌

[@VishnuTrinai]

Hi. Bro your videos are so helpful, i working on some ai projects and it needs web scraping,i have no idea . You posted a video. Im working on frontend you posted streamlit,now im working agents you posted this video.💥Your videos are so helpful. Thank you.🔥🔥

[@python.madeeasy]

hi Tim, thanks for your wonderful work. it would be nice if you could make full ai course video from beginners to advance

[@Tufelkind]

lmao you went from baby steps to like a 5k in 10 minutes; super well explained though, and covers a ton more ground than other videos/creators in this area.

[@jamesmikeedison2281]

Great Video, but vectorize gives error message of incompatible with embed input

[@Afzal3000]

Tim Dropping some fire content 🔥🔥🔥

[@sabbirhossan3499]

Please make a video on gpt hell 😢

[@javimolinero]

Awesome video. I have a question related to how the RAG component is utilized here. If I am not missing anything, this example uses all the notes in the collection when using ask_ai_func(), no matter the user_id asking the question, which means that it will consider notes created by different people. Is it possible to filter the notes by some key?

[@AnishDeshpande-i1o]

Learning this will improve my chances of getting a job. Right? I am upskilling, and I am really enjoying this stuff. But is this what companies are looking for when they hire Devs or PMs?

[@dtreezy]

experience

[@22combo]

Great project . But I'm getting issues at the end of the "ASKAI" seems to be with the call agent tool and last openai model connected to the advice output "Error building Component OpenAI: Connection error" & second with downloading langflow package in the editor. Others seem to not be able to download the package too.

[@kevintembobine5368]

Thank you ! Do you know any other good alternative to langflow ?

[@lazyraconteur8279]

Please help me out. Anyone who has built the flow, can you say why is the "Conditional Router" not working properly? Whenever the LLM gives an input of "Yes" or "No", it only runs the "No" component of it. The calculator and the "Yes" component never works. However if I remove the "Conditional Router" and simply use the Tool calling Agent and the other components, it works perfectly.

[@argl1995]

Hey Tim, I have to build a multiagent RAG for which I need your guidance. Is there a way to contact you apart from KZbin so that I can share the problem statement with you?

[@MythiliSoundhararajan]

Hello sir, may I know what the prerequisites i need to learn to do this project? ....

[@Epipedobatideo]

I'm not Tim (clearly), but python intermediate level (datatypes, loops, functions, etc) and knowledge about streamlit (Tim has a minicourse about it), seems to be sufficient to understand this project in my opinion. Knowledge about RAGs, Langflow etc helps but I wouldn't say required to understand (but to eventually implement is appropriate).

[@americanpandit9885]

Unable to use ASTRA Vectorize. It shows that it was DEPRECATED. Please help me resolve this issue.

[@saicharan6333]

after the recent update, ASTRA vectorize need not to be used. In ASTRA DB block itself there is an option "embedding Model or Astra vectorized" . choose astra vectorize then you will be able use it with NVIDIA embedding model

[@ANSTONDENVYLDSOUZA]

@@saicharan6333 only NVIDIA model? bcuz im currently using the Google gemini

[@Blimaxx]

Love you Tim

[@imperialgamer5376]

hey i am trying to build the same thing with groq api but i running into some issues i am able to execute it in langflow but when i try to run locally i get an error that there is no api_key even after mentioning in .env file

[@AjayCoding]

Great Vdo!!

[@ahmadzaimhilmi]

What's the bare minimum cost to hire you to set up one. I'm on a tight budget here.

[@TechWithTim]

$200/hour

[@harshal8508]

what is alternative to astradb.

[@gantumurnatsagdorj2306]

Error building Component Tool Calling Agent: Prompt must contain 'input' key. ?

[@jamesmikeedison2281]

add {input} inside the prompt field and it should work.

[@deepakkrishna2998]

Thank you

[@richardgriffiths3277]

Can anybody help? No matter how many times I enter the API key for the astra db I get an error? I have used just Astra DB component as Vectorize being deprecrated. "Error building component astra db: Error initiliazing AstraDBVectorStore: [401] Unauthorized Invalid JWT serialization: Missing dot delimeter(s) Please check or regenerate your API key."

[@staceyvanwyk6746]

Hi my model is not running can you maybe assist with fixing it

[@gagansingh3481]

Instead of using open ai api key is there any alternative to use model without api key

[@TanmayBhatgare]

you can use gemini flash 1.5 just log in with google developer and collect API key from there you need to have 18+ yrs account for this.

[@anigilajeyusuf8201]

Hugging face

[@jacksmith-ih9rm]

Great!!

[@TanmayBhatgare]

Hey Tim, I don't know is it only me or not but connecting to Tool Calling Agent to Text Response, it just popped up as, "No compatible components found", but on other hand you were able to make it connect idk how but how to solve this? 24:50

[@rohiteswargouthu6422]

did you solve it

[@TanmayBhatgare]

@rohiteswargouthu6422 yeah mate you just have to run langflow locally using "python -m langflow run" command. This is the only way which will work. Langflow website is still in initial phase

[@TanmayBhatgare]

​@@rohiteswargouthu6422 yeah I did. You just have to run langflow locally, with "python -m langflow run" and you'll be able to resolve this issue.

[@PANDURANG99]

Great

[@SkiLarkk1]

Hey Tim, or anyone else that can help, I'm at 24:51 and when I go to connect the tool calling agent response to the text output it just isn't connecting for whatever reason. Any idea as to what I can do? Thank you!

[@TanmayBhatgare]

I got the same issue, tool agent just say, "No compatible components found". You got any solution yet?

[@SkiLarkk1]

@@TanmayBhatgareno not yet unfortunately, I reached out to their team and asked in their discord, once I get it resolved I’ll be back to let you know

[@TanmayBhatgare]

@@SkiLarkk1 I too raised a GitHub issue, hope it will help.

[@cerrodelsol2668]

Me too. I have the same problems.

[@SkiLarkk1]

@@TanmayBhatgare its been insanely hard to get decent support for this lol

[@kichu2759]

I ran the project it works but i dont get the the expected output in Ask AI section with the details ive entered above but rather get the output based on what i've tested in LangFlow in Browser ( The quesiton and Answer ). Idk what to do! Can you help me out

[@TechWithTim]

You need to clear out the results from langflow first

[@jamesmikeedison2281]

Oh wow? How long did it take you to run pip install langflow. This is taking forever for me

[@kshitijjaiswal1945]

I am not able to connect the tool calling agent(response -> text) to output. it says no compatible components found (24:54)

[@TanmayBhatgare]

same issue here i think this is because it's in beta phase...you got any solution for this?

[@ANSTONDENVYLDSOUZA]

Im not able to access the playground at all. i hover over the playground button and it give me this message "Add a chat input or chat output to use the playground"

[@TechWithTim]

Ya you need to have an input and output in order to use it. Try swapping the text input with chat input and same for the output

[@ANSTONDENVYLDSOUZA]

@@TechWithTim some of the components are suddenly outdated. Like you wont find the conditional router anymore.

[@Mehfooj-i5l]

Anyone, Can I use llama api because OpenAI APi is paid ?

[@kichu2759]

Yes, You can!

[@betsonwilliam9470]

❤

[@faisalishfaqciiisilver275]

What do you mean by multi agent?

[@AaronBlox-h2t]

An agent is just an LLM that can use a tool aka another software program....yeah, tech jargon, alsways making stuff sound more complex haha

[@StynerDev-DominicanUniversity]

🎉🎉🎉🎉🎉🎉

[@nizardhahri493]

how can I link a local ollama model ?

[@deepakkrishna2998]

Did you figure out how to use ollama?

[@AaronBlox-h2t]

THIS videois only 2 weeks old and the project code is already broken??? That is no good....

[@TechWithTim]

How is it broken?

[@zombie-lwg]

is it all free

[@dabeluchi]

Very informative

[@Baboosh-z7u]

I'm unable to run the code locally as it's show ..user I'd is not set for AstraVectorStoreComponent.....help me out this