From an ex-Meta (Facebook) data scientist who used to build AI models for Instagram’s algorithms:
How To Get Ahead In Your Career with AI Agents
Watch the short video below to see how in the next 28 days you can learn one of the most in-demand AI skills in 2025.
Hi friend! How's it going??!
So if you're a business owner or professional who wants to get in their career by learning one of the most in-demand skills in 2025…
I’m going to show you how within 28 days, you’ll be building advanced multi-agent systems.
Systems that are reliable, produce high quality work and are valuable to any business.
You don't need prior coding knowledge. Just a desire to learn.
And you can stop wasting nights and weekends deciphering tutorials that leave you confused and disappointed.
Even with a busy schedule, this time next month, you won't be on the sidelines trying to 'keep up' with the latest AI news. You'll be at the forefront, actually implementing AI agent systems into your workflow.
Sorting your inbox…
Doing your research…
Finding business leads...
Managing your calendar...
All so you can focus on the tasks AI can't replace.
And that's just the beginning.
Not only will you be able to automate your own workflows.
You'll be able to automate workflows in yours or your employer's business. And you'll even be able to sell this as a freelance service if you choose to.
You'll be able to look at any business, spot the best areas to automate and create an agentic system faster than you ever imagined possible.
Right now, we're witnessing the greatest shift in how white-collar work gets done since the industrial revolution.
A divide is forming between those who can use AI and those who can't.
In 28 days time you can be in the tiny camp of users who not only use AI, but leverage it to the highest degree.
Before I share exactly how I can help you do this, let me make it clear.
This is NOT for you if:
You've already built multi-agent AI systems
You're resistant to using AI to automate workflows
You're an advanced programmer seeking deep technical knowledge
You have no interest in practical application and just want to learn the theory
You don't have at least 3 hours a week to invest in this
If you passed the first test, great!
I'm super excited to share this with you.
Now, Let me Introduce Myself… 👋
You probably know me from my YouTube channel or emails.
My name is Tina Huang, an ex-Meta (Facebook) data scientist with a 800K AI and Tech YouTube channel.
I've been working with AI for 7 years now.
My first project was with Google’s BERT at the University of Pennsylvania for my Computer Science Master’s in 2018.
(That was back when getting an AI model to speak a few lines of Shakespeare was considered groundbreaking….)
I also worked with AI for Instagram when I joined Meta in 2020.
We were building AI models for content policing algorithms, where our models would analyze content and automatically check for community guideline violations.
Then after leaving Meta to focus on YouTube…
I founded Lonely Octopus because I noticed a problem with most data science/AI boot camps and courses.
Not only were they about as engaging as school.
(For context, I did not enjoy school and honestly struggled to focus and learn in that environment.)
But they were either all in-person and impossible for working professionals to attend - especially if you have a family to look after.
Or they were online without any form of community-learning.
It’s no surprise that course completion rates are so low…
There wasn't anything that was personalized, engaging and had that feeling of accountability you get when you’re learning as a group.
So I launched Lonely Octopus.
A program designed to teach people data science and AI skills in the most engaging way.
We focus on the fundamentals and practical project building all alongside a community where everyone keeps each other accountable.
And over the last few years we've helped hundreds of people get ahead in their careers.
One of our most incredible case studies is Stephanie.
She’d been an English teacher for 14 years and was desperate for a career change..
But she didn't know where to begin gaining the credentials needed to impress employers.
So she joined Lonely Octopus in 2024 and after a lot of hard work is now in the AI team at one of the leading tech companies in the world, who've built one of the top AI models that you're likely using today.
Of course not everyone has ended up working for clients of this size.
But almost everyone has advanced in their careers and become the go-to AI person in their network.
Ok Let's Talk About AI Agents In 2025…
In 2025 agents have never been more powerful.
The better the AI models get, the more AI agents can do.
It's honestly crazy how much the software world is changing
As Bill Gates wrote, "agents are not only going to change how everyone interacts with computers. They're also going to upend the software industry, bringing about the biggest revolution in computing since we went from typing commands to tapping on icons."
Over the last year, I've been building a lot of AI agents - code and no-code.
Some have been for myself, like Inky, my no-code personal assistant. It helps combat my ADHD by planning and scheduling my day.
Others have been for businesses we've helped inside Lonely Octopus.
These have included agents for content generation, customer support, data reporting, researching and much more.
We even built a cybersecurity agent that reviewed data and provided insights and action items to help with threat detection for a cybersecurity platform.
All to say, it’s pretty incredible what you can now automate with AI.
Today, the average consumer has access to tech that was previously only available to companies like Meta.
And it looks like my audience has caught on to this.
I recently ran a survey in my newsletter and the most requested topic was AI agents.
So because of this and the increasing capabilities of AI agents…
Lonely Octopus is launching the AI Agent Bootcamp!
Introducing the AI Agent Bootcamp
The AI Agent Bootcamp gets you to build your first advanced multi-agent systems in just 28 days.
It's a hands-on program which means you learn by actually doing.
By the end you'll have built a total of 4 agent projects.
Starting simple and ending with an advanced multi-agent system.
Whether you can code or not, the bootcamp will allow you to actually code your agents.
You can stick to no-code if you prefer, but coded agents have their advantages.
They’re cheaper, easier to customize and allow you to have complete confidence in building systems for more complex workflows.
The Roadmap To Building Advanced Multi-Agent Systems In Just 28 days
Most people start building agents by diving straight in the how-to and copying a no-code template.
But they quickly hit a technical roadblock.
Due to the lack of fundamental knowledge, they end up with a half-working agent and wonder what all the hype is about.
But when you learn the fundamentals BEFORE the how-to…
You learn much faster.
And are able to build more advanced solutions due to your deeper understanding.
So each week of the bootcamp begins with a super engaging 60-minute masterclass.
It starts with the fundamental frameworks needed to complete the week's project.
Then ends with an agent demonstration so you can see the frameworks in action.
This is followed by a Q&A to answer all of your questions.
On Saturday, we meet again for a relaxed Q&A session.
You can ask for feedback or support on your projects or just sit and listen in the background. Many people treat this like a mini co-working session where you turn up, work on your project, and just ask a question if you get stuck.
You will also get lifetime access to all the recordings, so you don't even need to turn up to calls live and can just submit questions into the community.
Now let’s dive into what you’ll build each week…
Week 0: Laying the AI Agent Foundations
The first week is all about laying the foundations for you to build the best AI agents.
We’ll cover what an AI agent really is.
And dive into their true capabilities so you can pick the best projects to build.
Then we’ll help you set up your full agent tech-stack, map your project goals and show you how to get the most out of the bootcamp.
During the introduction week, you’ll learn:
The basic of agents and what they are are capable of
How to plan and scope projects so they’re completed quickly
Industry-level agent tech-stack so you can build anything from a simple agent to a robust multi-agent system interacting with the real world
What you’ll do during the week:
Setup your full AI Agent tech-stack using our setup checklist
Set your project goals, roadmap and milestones
Meet me, your mentors and peers
Now onto the exciting stuff…
Week 1: How to Build An AI Agent
Next, we’ll walk you through the fundamentals frameworks you need to build AI agents.
You’ll learn how to get the most out of each component - the prompts, models, tools, knowledge, memory and more.
And you’ll learn how an agent ‘thinks’ so you can more easily map out the best agentic system.
By the end of the week, you’ll know the key fundamentals and have built a simple that focuses on executing one task very well..
Some of what you’ll learn:
How to identify the best use-cases for AI agents
What most people miss about the anatomy of an AI agent
The #1 mistake businesses make when building AI agents
Prompt frameworks adapted from Google that allow your agents to output superior results
The difference between code and no-code agents and how to build either (even as a beginner)
What you’ll build during the week:
Scope a simple valuable agent use case
Building a simple AI agent
Example of what you’ll build:
Task Generator: Enter a goal (like “launch a website”) and the agent returns a structured to-do list broken into steps.
Now the basics are covered, let’s give your agent some power…
Week 2: How to Make An Agent ‘Human-Like’
It’s time to make your agent genuinely useful in the real world.
During this week, you’ll see how you make your agent ‘context-aware’ so that it can adapt and make decisions based on real-world data.
You’ll also build multi-step workflows and tools for your agent to execute so that it can perform more complex tasks.
Your agent will now start to feel like a mini co-worker who can automate parts of your workflow.
Some of what you’ll learn:
The principles of tool selection and architecture that most people skip (this inevitably hurts their agent performance down the line)
Prompt engineering principles for effective tool-chaining so your agent can handle more complex tasks
How to connect your agent to live data so it can react in real-time
A simple system design rule so your agent doesn’t fail from edge-case scenarios
How to add context-awareness to your agent to make it more human-like
What you’ll achieve during the week:
Build a context-aware agent that interacts with external information.
Example of what you’ll build:
Research Agent: Accepts a query like “AI trends in healthcare,” performs a web search, and returns a detailed summary with links
Policy Bot: Allows users to ask questions like “What’s the PTO policy?” and retrieves answers from an internal handbook or PDF
CRM Helper: Looks up contact records in a database or spreadsheet and returns the latest client details or notes
Now for the final phase of agentic system building…
Week 3: Build An Advanced Multi-Agent System
During this week, you’ll build your first ‘team’ of AI agents.
You’ll have an agentic system built of individual agents each with their own role, communicating with each other seamlessly to automate an entire workflow.
We’ll start by covering key multi-agent design patterns.
You’ll learn the strengths and weaknesses of each one so you know the perfect one to apply to your projects.
This is when all the previous work you’ve done will really start to click and connect together.
Some of what you’ll learn:
The best business use-cases for multi-agent systems
How to pick the best system design patterns for you project
How to give your agents memory so that they become more intelligent
How to use guardrails and fallback logic so you system holds up in any scenario
The best way to get your agents to seamlessly communicate with each other
What you’ll achieve during the week:
Build your first advanced multi-agent system
Example of what you’ll build:
Content Generator: One agent finds references, another outlines an article, and a third drafts content, all in sequence.
Lead Qualifier: Accepts new inbound leads, classifies them by type, and routes each to the appropriate email or follow-up path.
Hiring Assistant: Screens resumes, compares candidates to job descriptions, and books top candidates for interviews
And that’s not all….
In addition to the above, you’ll also get:
In Addition to This…
Every week is accompanied by a 15 to 30 minute step-by-step pre-recorded tutorial.
There's a code and no-code version so you can choose either or even do both.
The tutorial walks you through how to actually use the tools with a live build of each type of agent.
And that’s not all…
As this is the launch of our first ever AI Agent Bootcamp, we’re including a few additional bonuses.
Bonus #1: How to Launch Your Agent to the Public
Although you’re now built a multi-agent system and are able to build many more.
We want to take you a step further.
So we’re including an additional bonus week where you actually deploy your AI agent so that it’s usable by the general public.
You’ll learn some advanced debugging techniques to make sure your agent is robust enough to deploy.
And you’ll learn to package your agent so the public can easily interact with it.
We’ll also cover some storytelling techniques to help position and pitch your agents so that people and businesses want to use them.
Some of what you’ll learn:
UI essentials that immediately improve the user flow of your agents
How to trace agent behaviour so you can quickly and easily find and resolve bugs
The easiest way to package and deploy agents (code and no-code) so that they can be used by the public
Storytelling secrets every AI builder needs to know so that you can position and pitch your agents as valuable solutions
What you’ll build:
A full pitch and demo of your multi-agent system (optional)
Build and deploy multi-agent system that’s polished and ready for public use.
Examples of what you’ll build:
Personal Financial Advisor: Actively analyzes your spending data to finds patterns, generator reports and suggests saving strategies or investment ideas
Document Processing Pipeline: It extracts key data from your internal files (notes, invoices, reports), classifies it then summarizes or routes the output - simulating a real B2B automation workflow.
Interview Coach: Offers personalized job prep based on resumes and job descriptions.It generates practice questions, evaluates your answers, finds your gaps and suggests improvements
Bonus #2: An Introduction to LLMs by Andy Levenson
Last year I hosted Andy Levenson for an incredible private workshop.
Andy is a data scientist specializing in natural language processing (NLP).
He has over 10 years of experience and currently leads a data science and analytics team at a Fortune 500 company.
By the end of this workshop, you’ll understand exactly what goes on under the hood of an LLM.
You’ll really ‘get’ how an AI model thinks, which will change the way you interact and improve all of your AI outputs moving forward.
Some of what’s covered in this workshop on LLMs he covers:
The real dangers we should worry about instead of "AI extinction"
Everything you need to know about how ChatGPT actually works under the hood
How to get dramatically better results from AI by using this simple prompting technique
Why the shift to "Transformer" models in 2017 created the AI explosion we're seeing today
Why even simple chatbots from the 1960s created parasocial relationships and how today's models pose real psychological risks
Now onto the final bonus…
Bonus #3: How to Study Technical Things
Earlier this year, I hosted a private workshop on how to study technical things.
When you finish watching it, you’ll know the best self-learning techniques and how to use AI to amplify them further.
You’ll feel 10x more confident learning any topic, no matter how technical.
And you’ll learn how to maintain motivation for learning, even when you’re busy in your regular life.
Some of what’s covered:
The hidden "80/20 rule" for learning technical topics faster
How "accountability hacking" stops you from ever giving up
Why your current note-taking method is slowing down your learning speed
The "priming technique" that top learners use to absorb information more effectively
The "lagging vs. leading metrics" framework that prevents burnout
5 specific prompt techniques to turn AI into your personal coding tutor
Now, here’s some words from people who’ve joined Lonely Octopus program’s in the past.
“I'm a data analyst / product manager working in edtech and joined Lonely Octopus back in 2024. Lonely Octopus really helped me seize the momentum of AI/LLM. Since then I have changed jobs and now work at my dream job (helping develop AI products for learning). I highly recommend LO not only because of the team's technical expertise in AI but also because of their genuine care and support for your success!”
— Jinjing, Data Analyst
“I'm a software engineer who attended Lonely Octopus in the summer of 2024. Opportunities have risen due to my experience in AI from Lonely Octopus and I couldn't be more grateful to their team and knowledgeable instructors. As I've completely transitioned to a career in AI Software Engineering, it's been a direct causation due to my time with the cohort. So for anyone looking to advance their AI skills or learn an ever evolving field for the first time, Lonely Octopus is absolute professionalism and they provide the support necessary to truly take AI into your own hands.”
— Tyler, Software Engineer
“Absolutely loving this experience! You all created a wonderful little world of learning, and I am so impressed with not only the content but the environment you've created.”
— Shelby, AI Applications Designer
“One thing that really surprised me about Lonely Octopus is its personalized touch and that they really care about you and are willing to help you to reach your goals.”
— Jeshua, Data Scientist & Analyst
- Week 0: Laying the AI agent foundations
- Week 1: How to build an AI agent
- Week 2: How to make an agent ‘human-like’
- Week 3: How to build an advanced multi-agent system
- No-code and code tutorials for all the tools and projects
- Weekly Q+A session every Saturday to get live feedback and support on your projects
- Community access: Build AI agents alongside 100 other peers on the same path as you
- Daily support: Ask us questions in the community anytime you need assistance or feedback
- Bonus #1: How to deploy your AI agent to the public
- Bonus #2: An introduction to LLMs
- Bonus #3: How to study technical things
- Lifetime access to all recordings
- 14-day money-back guarantee - no questions asked
What’s The Investment?
- Everything included in AI Agent Core
- 2 × group mastermind calls with Tina
- 5 × 1-on-1 calls with a mentor during the bootcamp
- 3 × additional consulting calls to be used within 6 months
- 14-day money-back guarantee - no questions asked
P.S. We have to limit spots so that you get all the attention you need. This way you’ll never have a question go unanswered. 😊
Tina’s AI Agent Promise
Although I believe you’ll be amazed at how far your AI skills progress with the community style learning and the focus on fundamentals.
There’s an off chance the AI Agent Bootcamp isn’t what you expected or you realize building AI agents just isn’t for you.
If this is the case, you can get a full refund anytime within the first 14 days.
Just drop an email to contact@lonelyoctopus.com and you’ll have your money back in just 48 hours.
This needs to be a positive return on investment for both of us. 😊
Your Questions Answered
Will I really have a working team of AI agents?
Absolutely! By the end of our 28-day bootcamp, you'll have built 4 complete AI agent projects, including a full multi-agent team. You'll start with simple agents in Week 1 and gradually build up to more complex systems. And unlike other programs that just teach theory, our hands-on approach means you'll actually have working agents that can sort your inbox, do research, manage your calendar, and find business leads - all working for you 24/7!
Will building AI agents still be a valuable skill in the future?
It might not be clear exactly which form AI agents will be in the distant future, but they are here to stick around one way or the other.
As Bill Gates said, “agents are not only going to change how everyone interacts with computers. They’re also going to upend the software industry, bringing about the biggest revolution in computing since we went from typing commands to tapping on icons.“
How do I know if this is for me?
If you're a professional or business owner who wants to get ahead in your career by building AI agents and you don’t fit into any of these:
You’ve already build multi-agent AI systems
You’re resistant to using AI to automate workflows
You have no interested in practical application and just want to learn the theory
You’re an advanced programmer seeking deep technical machine learning knowledge
You don't have at least 3 hours a week to invest in this
…then this bootcamp is for you!
You don’t need any prior coding knowledge. Just a desire to learn and to be excited about automating your tasks with AI.
How is this bootcamp different from other options?
Firstly, I’m not like many AI gurus who just hopped on the ChatGPT train the moment it came out. Me and my team had deep knowledge and experience around machine learning, data and AI before ChatGPT came out. You’ll be learning about AI agent systems on a deep and fundamental level.
The bootcamp is also focused on you actually building projects and doing this alongside a fun community to keep everyone accountable.
When can I expect to see results?
Within the 28 days you’ll have a full working multi-agent team! And even before that, by week two you’ll have your first working AI agent. This doesn’t factor in the fundamental skills and knowledge you’ll be picking up from day 1.
What if I decide it isn't right for me?
On the off-chance you don’t feel like it’s for you, simply email contact@lonelyoctopus.com within the first 14 days and it will be processed within 24 to 48 hours - no questions asked.
This needs to be a positive ROI for both of us. 😊
“I currently work on the data team for a major AI company. Before that time, I was an English teacher and desperately wanted a career change but didn't know where to begin gaining the credentials needed to start impressing employers. Since I was coming from a non-STEM field and trying to enter STEM, my biggest challenge was just starting at the bottom of a whole new ladder. Lonely Octopus gave me exactly the coursework foundation and practical projects I needed to start totally reshaping my resume.”
— Stephanie, Data scientist, AI Trainer and Prompt Engineer
“The Lonely Octopus team has been great by always being available to give direction, advice, and support.”
— April, Cybersecurity & Compliance Professional
“The community continues to be a tremendous source of inspiration and motivation for me.”
— Bhakti, Analytics Engineer
“The program has created a rich community of learners from various backgrounds.”
— Jihui, Analytical Scientist
“The community was incredibly interactive, creating such an immersive and engaging environment. The instructors were also consistently responsive and available, which made the experience even more valuable.”
— Jacky, Software Engineer
“The office hours have been invaluable to my learning experience. It’s incredibly helpful to have someone available to discuss topics and clarify any confusion. ”
— Donald, Warehouse Developer
“I'm just super grateful for this opportunity. The Lonely Octopus program has already pushed me way beyond my comfort zone.”
— Fernando, Data Analyst
