16 Months of Building AI Agents in 60 Minutes

Here I go through a core overview of information that I have learned through the past 16 months of building AI Systems in the form of AI Automations and AI Agents, both for my agency as well as for my AI startup. All of this knowledge in 60 minutes. I tried to cut out all the BS and get straight to the point. I discuss fundamental concepts that you should understand, in order to harness LLMs in the most effective way. These points of view, ideas and concepts should allow you to speed up your knowledge up to date and create better solutions for yourselves and for your clients, as well as understanding the potential and future of Agentic Systems.

There a typo on the slide at 31:32; it's an Airbnb agent not a chatbot since it has access to tools.

I also briefly go through n8n to show you how the platform allows you to visualise and use AI Agents effectively. Not saying that Make is a bad tool, just letting you know that n8n is another tool that you should explore.

IMPORTANT: Let me know if you see something that you would like me to make a continuation video on. This will allow me to understand where I can help you the most.

🥣 RESOURCES:

🔹 EXCLUSIVE COMMUNITY:
I am receiving many DMs about building a private, exclusive community where I will give you access to automations never seen before yet and a behind the scenes of the workings of my businesses, amongst other things. Before even doing that that I want to check interest by allowing you to quickly fill in this form, thank you! :

🔗 VISIT FGLABS:

FOLLOW DOCUSPHERE ON LINKEDIN:
We are changing the way people interact with data. Follow us on LinkedIn to stay updated with the latest changes:

Time Stamps:
0:00 Intro
2:41 Back to the Basics; what exactly is an LLM?
6:02 How LangChain enabled the new wave of AI Applications
6:56 Where LangChain lies in the Abstraction Layers
10:09 LangChain Basics
13:05 Clarifying what an AI Agent is
16:42 Understanding RAG Systems
21:12 Multi Agentic Systems
28:15 The difference between AI Agents and AI Automations
31:18 Building your First Agent
33:32 Little Demo- Airbnb Agent on n8n
43:46 Adoption of Agents in Enterprise
45:16 Tips and Considerations for building AI Systems
59:06 Agents are the Future

Please like and subscribe if you found this video helpful in some way.

#ai #aiagents #aiautomation #openai #makemoneyonline #n8n #make

[JayeeeMaooo]

Great Video. I have used exactly same way building agents for about 1 year. several things to add:

1. When design automations, scalability, robustness, easy-maintained and error handling are top criterias when hand in to the client-side.

2. Human loop decisions and judgement and approvals matters.

3. LLM request body parameters(like temperature....) is a nice to have, but in advanced cases, it's a must have.

4.front-end interface (n8n have forms and html template to start with)

5. vertical domain logic design and operations is always ignored among all AI builders, talk to experts and get more vertical knowledge to optimize workflow and prompts.

[funnydogfargo]

A whole video about WhatsApp will be great. I've built an AI agent workflow as a backend, created WhatsApp Business Acc, and set it all up, but can't figure out how to connect the front-end (WhatsApp) with the back-end (AI Agent workflow). That way I choose Telegram instead. Slack is also easy, but in my Region (North EU) Slack is not a common app for business.

[ismile]

Thank you so much. Covered end to end with more details. It's easily understable, and you made the thing simple and interesting to match 1 hour video by going step by step. Just want to know where we can download ppt? Is this sample available anyware to downlaod and test?

[giangallegos]

Another epic tutorial!!! I want to learn more about the workflows!!

[TaohidKhan-utgf]

After watching only 5 minutes I can say this is amazing.

[WillMoody-crmstorm]

Thanks and yes, please. A whatsapp chatbot video would be fantastic!

[MinaEllis-dn]

Awesome! You hit it big again, you really presented your video in a wonderful and beautiful way. I love this and I love you too. You are so creative. This is the first video I have seen and I have already decided to subscribe!

[totlenmoingaycungban]

Super thanks your video. Well explanation with visual. AI Agent will change the game. We can create our tools to perform tasks more effectively. Those who with have "lazy mindset" and basic programming techniques can automate their tasks.

[nanavareyogesh]

Very well explained. I learnt a lot through this. loved it. Looking forward for videos on architecture designs.

[benoitcorvol]

Thank's for the video, it will be nice to have a video of how to construct a workflow of multiple ai-agent with different tool on autogen, maybe wait the 3.0 because i think it's coming soon :)

[bloodywolftr]

This is a very good overview of the possibilities that awaits there. I loved the structure of the video and depth of it.Thank you very much for taking your time to make this and sharing with us. The video aspect ratio for the slide doesn't fit to 16/9 format. I am interested in automation in local environment with local llms without using any api, web services etc first. But of course with rag live data retrieval + evaluation on that data with multiple agents or llms sounds really nice. However I am a bit lost about where to start. Is it possible for you to start a tutorial series for complete beginners on this stuff? Especially on local environment first if possible.

[ward_jl]

Great video man, clear way of explaining agents. Start with the fundamentals, to more complex agent structures. Also, it is nice to hear your view on these systems' future (market) potential!

[pohkeamtan]

I enjoyed this video very much. You live up to your promiseto " I tried to cut out all the BS and get straight to the point."

[germanleyva]

Amazing video about Ai agents (the future) a video on how to make voice agents would be great! a chatbot Honestly, I'll consume any information you provide because it is all connected at the end of the day! Thank you!

[meelijah]

Hey Frank, thank you so much for the great content. Please set up a community, I will be the first paying member.

[jonk]

Frank - This is great and helpful. Could you go over
1) the business side of setting up a similar agency as yours, such as pricing strategy, delivery, requirements gathering, etc.?, and
2) how to build these multi-agents using pure code and deploying it ourselves, instead of through a variety of nocode platforms?

[charmsoflove]

Thank you great video. Would love to see one for architectures and voice agents. ❤🎉

[samben]

Great presentation Frank. Will you share the google doc for the prompt example please? Thank you.

[theintern]

Wow this was amazing as a beginner who wants to learn more

[sossassessissus]

Hey! Nice video.

Do you think it's better to use N8N than direct python programming even if im software engineer? (i have no problem programming anything and have exprience coding in much languajes)

Like, do you think it has any advantage for those who already knows how to program and tried CrewAI and stuff like that?