Agentic Machine Learning Engineering Bootcamp

Start by setting up your professional data science environment, master command-line and Git workflows, and refresh your Python skills for data manipulation and visualization. You will then apply these fundamentals to conduct an end-to-end exploratory data analysis on a real dataset.

Build a strong foundation in AI and machine learning by applying classification and regression techniques using standard ML workflows. Learn to evaluate and optimize models with industry-standard metrics, hyperparameter tuning, feature engineering, regularization, and cross-validation, while gaining an understanding of gradient descent and writing reusable Python functions.

Advance your skills with decision tree ensembles, boosting, and techniques for handling class imbalance, then explore clustering and dimensionality reduction to solve real business problems.

You'll explore deep learning through artificial and convolutional neural networks, learning how backpropagation extends gradient descent to train deep models. Then, you'll build and train neural networks using an industry-standard deep learning framework and apply transfer learning to image classification tasks.

Move from classical machine learning to modern large language models by understanding transformer architectures and the foundations behind LLMs. Gain hands-on experience with LLM APIs and prompt engineering, build an end-to-end Retrieval-Augmented Generation (RAG) system, and learn when and how to fine-tune models.

You’ll build agents with persistent knowledge and memory, mastering agentic RAG patterns, adaptive retrieval routing, and memory architectures from short-term context to graph-based long-term recall.

Transition from notebook-style experimentation to structured GitHub repositories as you uncover the foundations of agentic AI. You will implement the core agent loop from first principles, refactor your code to an industry-standard agent framework, and ship your first end-to-end agent equipped with sandboxed tools.

Next, you'll learn stateful, durable, and debuggable workflows. You'll implement streaming, human-in-the-loop approvals with strict idempotency, and instrument complete, production-grade observability across your entire system.

You’ll achieve fluency across the modern agent stack: consuming and publishing agent tool protocols for external integrations, working with autonomous-tool-use frameworks (including hook and subagent patterns), and applying type-safe, contract-driven design to production agents.

Finally, you’ll orchestrate multi-agent systems using the optimal pattern for the task, while fine-tuning the architecture to balance cost and latency.

You’ll ship production-grade agents with rigorous evaluation and defense-in-depth safety, and gain a working awareness of AI governance and EU AI Act obligations.

The capstone begins as you frame your project, design its architecture, and build an initial end-to-end implementation with tool integrations, knowledge retrieval, and memory. Next, you’ll refine your system by iterating against an evaluation suite and presenting your progress for feedback. Finally, you’ll harden the agent with defense-in-depth safety, deploy it to production with observability dashboards, and showcase your working system in a live graduation presentation.