Episode 2- How large language models (LLMs) actually work (at a high level)

Episode 5- Planning and multi-step reasoning in agents

Most AI systems don’t actually “plan.” They respond. There’s a difference. In this episode, we break down what planning and multi-step reasoning in AI agents really mean. You’ll learn how agents take a high-level goal and turn it into structured, ordered actions instead of a single answer. We explain how goals are decomposed, how decisions are made step by step, and how feedback loops help agents adjust when things change. You’ll also understand where failures usually happen and why execution control matters as much as model intelligence. By the end, you’ll be able to evaluate agent designs, scope realistic use cases, and decide when multi-step autonomy is worth the cost and complexity.

Episode 4- Memory in AI agents (short-term vs long-term)

Most AI systems don’t actually “remember” anything. Unless memory is deliberately engineered, every interaction starts from zero. In this episode, we break down memory in AI agents, specifically the difference between short-term memory inside a model’s context window and long-term memory stored externally in databases or vector stores. You’ll understand what each type does and why both are necessary for real-world agents. We explain how memory is stored, retrieved, and reinserted into the model’s reasoning process, and where systems typically fail. By the end, you’ll be able to evaluate agent architectures, design for persistence, and make informed product decisions around personalization, continuity, and cost.

Episode 3- What “tools” and API calling mean in agent systems

Text isn’t enough to get real work done. In this episode, we explain what “tools” and API calling actually mean inside AI agent systems, and why they’re the difference between a model that talks and a system that acts. You’ll learn how agents connect to external software like CRMs, databases, email, payments, and internal systems, how APIs let them send and receive structured data, and how this turns language into real-world execution. We cover the basics without jargon: what an API is, how tool calls work step by step, common failure points, and the reliability and security tradeoffs teams face in production. If you want to build or evaluate real agents, this is the foundation that makes everything else possible.

Episode 2- How large language models (LLMs) actually work (at a high level)

LLMs feel intelligent. But under the hood, they’re not thinking. They’re predicting. In this episode, we explain in plain English how large language models like ChatGPT actually work. What they’re trained on, what a “token” really is, how probabilities drive every word they generate, and why they sometimes sound confident but still get things wrong. We break down the core mechanics step by step: training on massive text data, learning patterns, predicting the next word, and using context to stay coherent. No math. No jargon. Just the mental model you need to reason clearly about what these systems can and cannot do. By the end, you’ll understand their strengths, limits, and how to use them wisely in real products and workflows.

Episode 1- What is an AI agent vs a chatbot?

Everyone says they’re building “AI agents.” Most are still just chatbots. In this episode, we break down the real difference between chatbots, large language models (LLMs), and true AI agents that can plan, use tools, and execute multi-step tasks across systems. You’ll learn how agents actually work under the hood, where they fail, and when you should use a simple chatbot instead. We cover real business examples, architecture basics, risks, and how to evaluate products or vendors without getting fooled by marketing. If you design products, run operations, or make build vs buy decisions, this is the practical foundation you need.