
Agentic AI Essentials
AI isn't just answering questions anymore — it's setting goals, making plans, and taking action on its own.
Starts Tue, Oct 13
11:30 PM UTC · 1h 30m
4 sessions

Lloyd
Outcomes, not lecture notes.
Explain how agentic AI systems differ from traditional prompt-response tools and why that distinction matters
Define clear goals and scope for an AI agent so it stays focused, useful, and safe
Experiment hands-on with platforms like AutoGPT, CrewAI, Cognosys, and AgentOps to complete real multi-step tasks
Design multi-step AI workflows that operate semi-independently to save time and scale effort
Set guardrails and evaluate agent performance to catch errors before they compound
Identify appropriate use cases for agentic AI versus simpler AI tools
Prototype an agentic workflow applied to a real-world task or project of your choosing
Meet Lloyd.

Lloyd
Lloyd is an AI-native entrepreneur who specializes in automating businesses and processes with agentic workflows. His tech career started as "The YouTube Ads Guy," and when the early releases of ChatGPT and other LLMs emerged he saw AI as the most transformative technology of our lifetimes - and set about learning everything he could about it. Nowadays he's happy to check in between classes on the AI agents running his businesses while he shows Varsity Tutors students how they can do the same.
AI isn't just answering questions anymore — it's setting goals, making plans, and taking action on its own. Agentic AI Essentials is a four-session course that pulls back the curtain on this next wave of artificial intelligence: autonomous tools and multi-step workflows that don't just respond to a prompt but actually work toward an outcome. Students and forward-thinking learners will get hands-on with real platforms — including ChatGPT with memory, AutoGPT, CrewAI, and AgentOps — learning how to design, direct, and deploy AI agents for meaningful tasks. This isn't a survey of buzzwords; it's a practical, skills-first introduction to the technology that's already changing how professionals work, build, and solve problems. Whether you're curious about where AI is headed or ready to start building smarter workflows right now, this course gives you the framework to do it confidently and responsibly.
From Single Prompts to Autonomous Action
Most people's experience with AI looks like a conversation: you ask, it answers, you move on. Agentic AI breaks that mold entirely. Instead of waiting for your next instruction, agentic systems can set sub-goals, make decisions, use tools, and execute multi-step plans — all in pursuit of a larger objective you define. Agentic AI Essentials opens by grounding students in exactly what makes this new generation of AI different: how agents reason, how they chain actions together, and why this shift from reactive to autonomous represents one of the biggest leaps in how humans and machines collaborate. Students leave this first unit with a clear mental model they'll build on throughout the course.
Getting Hands-On with Real Tools
This course doesn't stay theoretical for long. Students will experiment directly with the platforms driving the agentic AI landscape today, including:
- ChatGPT with memory — understanding how persistent context changes what an AI can do for you over time
- AutoGPT — one of the original autonomous agent frameworks, used to explore how AI breaks big goals into executable steps
- CrewAI and Cognosys — newer platforms built for coordinating multiple AI agents working in parallel toward a shared goal
- AgentOps — a tool for monitoring, evaluating, and debugging agent performance in real workflows
Rather than passive demos, students are actively setting up tasks, watching agents work, and analyzing what goes right — and wrong.
Designing Agents That Actually Work
Building an effective agentic workflow is less about coding and more about thinking clearly: What is the goal? What does the agent need to know? Where does it need guardrails? This course dedicates serious time to the craft of agent design — how to write a task definition that gives an AI agent enough direction without over-constraining it, how to scope a workflow so it stays manageable, and how to evaluate whether an agent is genuinely completing your objective or just looking like it is. These are the skills that separate someone who tinkers with AI from someone who deploys it effectively.
Safety, Scope, and Knowing When to Intervene
Autonomy is powerful — and that's exactly why it requires guardrails. One of the course's most important threads is responsible deployment: what happens when an agent goes off-script, how to build in checkpoints and limits, and how to maintain meaningful human oversight in a semi-autonomous workflow. Students won't just learn how to let AI run; they'll learn when to let it run, when to constrain it, and how to evaluate its outputs critically. In a world where agentic tools are being handed more and more responsibility, this judgment is genuinely valuable.
What You'll Walk Away With
By the end of Agentic AI Essentials, students will have a concrete, working understanding of how autonomous AI systems operate — not as magic, but as engineered workflows with real logic and real limitations. More importantly, they'll have hands-on experience designing and testing those workflows themselves. Whether the goal is building smarter personal productivity systems, prototyping a tech project, or simply staying ahead of a fast-moving field, this course delivers the fluency to engage with agentic AI as a creator, not just a consumer.
Live Q&A
Cameras / mics optional
Recordings
Available within 1 hour, kept 90 days
Materials
No special materials required
Similar classes
free with a Varsity Tutors membership
36
DAYS19
HRS52
MINTue, Oct 13 • 11:30pm UTC
4 classes, starts in 36 days
Mon, Nov 23 • 10:30pm UTC
4 classes, starts in 77 days
Sat, Sep 12 • 2:00pm UTC
3 classes, starts in 5 days
Agentic AI Essentials

