Відео

Great video! Thanks so much for sharing. What tools do you use to record your videos and create beautiful visual animated diagrams?

Great content as always! You have shaped my AI journey through your videos bro!

Great work buddy!!

Great job, bro!

Amazing work!!! keep up the good work!

Amazing! God Job!

You are making really great contributions to the community, loved it

Amazing..Thank you bro

I run 12B model Q4_K_M on my 8GB card and you say we need 48GB of ram for 13B is insane. It is just waste of money.

very very nice, great work

hello bro... how can i add it to a website where people login and ask their queries

I built something like this some time ago. Great work. This channel will take off 2025. Postive.

¡Gran video, tío!

Excellent series - really useful to understand the Agent technology.

Very informative content, I am about to design a LinkedIn scrapper agent that will scrape the given profile, and gather their data in a csv and upon request sends them a connection request with personalized connection note as well..

0:57 this is no code.

Hello, did you try the notebooks in fullstackdeeplearning course? They seem not to be working for me.\

Can u create AI agents swarms project That you can use for content generation For example AI article website

it is NOT working for free, no GROQ possible with free codegpt subscription

Top. Thanks 😁

were you in constant communication with the ai? Could you then ask the ai to change the speed of the blink of just the red led? Im assuming that was the point of building this whole damn thing.

awesome. at which step are we storing the images in GCS?

Thank you so much. I have a question. started following you on youtube. have this question. I want to use agents for storytelling, and I have prepared templates for the stories to create standard/consistent stories (imagine a writer_agent and an editor_agent, etc). So, each agent is clearly defined, and each should use its own template as guidance to provide the output in the format I want. How do I have, for example, a writer agent use the writer_template to guide it on how I want it to write the story? The template could be part of the agent python code or a local txt file (I still ignore which option would make the agent output more efficient). Please, help.

Awesome vid, just like Andy says, like a boss !!

Love the content 🔥

Thank you so much! Your notebook and explanation is amazing! This is very important educational resources for the community! Really helped me to learn the fundamentals of LLM agent!!!

Thanks a lot for your content and explanations! It really helped me understand better how the agentic architecture works. Wishing success in your journey!

Thanks for the references, really great content on the channel ❤

Hi Miguel, congrats on the channel, keep doing these your videos are f* awesome :) Just a question , I tried building the Dockerfile.suno like you mentioned , but it always gets stuck in the "RUN npm install" command. How can i debug this? Ive looked for a similiar problem on the web but cant find a solution...Im using docker desktop on windows ( dont know if thats relevant or not because I dont really know docker that well). Abrazos desde Portugal ,hermano 💪💪

In my ReAct multiple agent system, the time it takes to obtain the final answer using tools can be quite long, leading to a poor user experience. In your system, how long does the process of calling the choose tool -> execute tool -> get final answer take in each round,, and how many times does the LLM (Large Language Model) need to be called?

Thanks for sharing your work Miguel - big fan of your content across your various channels. Interested in why you define the tool signatures in the system prompt as opposed to passing signatures as a list of tools to during the inference request. Have you noticed a difference in output doing it this way?

Great explanation, thank you

Nice. I need to do some tweaking to the agents, but this is awesome! Thank you

Banger video! Awesome resources. As a senior ML engineer, i especially endorse the Designing ML Systems Book from Chip Huyen

Thanka for the video ❤. Really appreciated ❤

Could you tell us how to pass an interview ?! Or what are the topic that must cover

What about math and basic stats and traditional ml.

The information shared here is gold ✨

I implemented you code in GCP using Gemini pro, I find Gemini really struggles on following the instruction of the prompt and understanding that it can use the observation returned to answer the question. You have to tell it specifically in great details in each tool's signature with "you can use this tool to do what, to do what, to do what"

Loved the video, man! So excited about our future collab.

Can i ask you something: why not break it into 2 steps: let llm first select a tool from the list of options(just define them simply and treat as a classification problem) and then in the second step get the values of the parameters. I have seen that when many tools are provided and we expect a complete executable function as output llms sometimes don't work? Also can we connect with you/follow you on linkedin, if it is okay with you? I have been trying to build things from scratch, although not as good as you do. Thanks!

The video is really good. You could actually use composio for tool calling and it would have made agent a lot more powerful.

enorme Miguel!! 🙌🙌

Very informative... Thanks for the good content...I do not understand people... I can see 9 views and 0 likes!!! ... although the content you provide is very special and truly good content. What you provided here is very nice and I liked the idea of applying topological sorting .. Keep the good work and have a nice day 🌹🌹🌹

🚀

Great content, thank you! Question for you, what tool did you use to draw your diagrams, those are rich and love the fonts you used.

Great Explanation! Welldone.

Simple, clear and to the point. Great video 🔥 Your cat wasn't that surprised by your videos 😂

Hey, waiting eagerly for the Multiagent Pattern. Please, soon?

Thanks for the good content 🌹🌹🌹