The AI Engineering Bootcamp

The course lasts 10 weeks, with content divided into weekly modules, each dedicated to a key topic in AI Engineering. The program gradually transitions from basics to advanced topics:

• Introduction and RAG: Week 1 introduces participants to the course (Introduction & Vibe Check), followed by studying embeddings and the Retrieval-Augmented Generation (RAG) method - searching for and using external knowledge to improve LLM responses.
• Applying RAG and LangGraph: Week 2 focuses on industrial use cases of RAG and creating end-to-end LLM pipelines. Students learn the LangGraph framework (a modern implementation of LangChain ideas) for building production systems using RAG.
• LLM Agents and Multi-Agent Systems: In Week 3, the concept of agent-based systems powered by LLM is introduced. Participants learn to create LLM agents using LangGraph and develop multi-agent applications where multiple agent modules interact with each other.
• Data Generation and Model Evaluation: Week 4 emphasizes generating synthetic data for testing and methods for evaluating LLM solutions. Students discover how to automatically generate datasets for model response verification and assess RAG pipelines and LLM agents based on metrics.
• Advanced Techniques: Week 5 covers in-depth strategies for search and information retrieval for RAG applications, as well as advanced reasoning techniques for LLM agents (to enhance their ability to plan and execute complex action chains).
• New Tools and Certification: During Weeks 6-7, participants undergo a special Certification Challenge (a practical test to verify acquired skills) and learn about the latest tools for AI engineers. The program introduces new frameworks and capabilities, such as the OpenAI Agents SDK, libraries for creating "code agents" (like Smol Agents), and other current technologies for working with LLMs.
• Deployment and LLM Ops: Week 8 is dedicated to launching applications into production. Participants learn to deploy LLM services via API, create production endpoints for their models, and get acquainted with the basics of LLM Ops - maintaining and monitoring LLM systems (e.g., logging requests, quality monitoring, model version management).
• Enterprise Practices: Week 9 examines corporate aspects of AI implementation. Topics include deploying LLM models in a company's internal infrastructure (on-premises solutions) and implementing mechanisms such as caching, versioning, limitations, and guardrails to ensure reliable operation of AI applications on an enterprise scale.

During Bootcamp, participants acquire a range of practical AI Engineering skills:

• Designing LLM Applications: The ability to develop architecture for programs using large language models - from chatbots and document search engines to complex multi-agent services. The course teaches understanding which components are needed to build reliable AI applications and how they are integrated.
• Working with Vector Embeddings and RAG: The skill to use data representation methods in vector form (embeddings) and implement the Retrieval-Augmented Generation approach. Graduates will be able to set up search across an external knowledge base and connect it to an LLM to provide relevant answers supported by data.
• Creating and Managing LLM Agents: The ability to develop autonomous AI agents that can perform tasks, requesting LLM models for decision-making. Students learn to use frameworks like LangChain/LangGraph to build agents and even entire multi-agent systems where multiple models/agents interact with each other.
• Evaluating and Debugging AI Systems: Experience applying LLM testing methodologies. Specifically, generating synthetic test data, automated comparison of model responses with benchmarks, and metrics for assessing model accuracy and consistency. These skills help improve the quality of LLM models and pipelines based on feedback and experimentation.
• Production and Support of AI Solutions: Practical skills in deploying AI projects in real conditions. Graduates will be able to deploy their models and services on cloud infrastructure, set up API endpoints for access, and implement client-side applications. Additionally, they will learn how to maintain the operability of such systems: implementing request caching, monitoring, limiting undesirable requests, and other MLOps/LLMOps elements for reliable model exploitation.
• Mastering Modern AI Development Tools: Throughout the course, participants master popular libraries and platforms used in the industry. They will gain skills in working with major model APIs (such as the OpenAI API), sectors of the Hugging Face ecosystem, frameworks like LangChain/LangGraph for building query chains, and rapid prototyping and deployment tools (like Vercel for web interfaces). The program is constantly updated to include the latest technologies: the latest course version includes, for example, the OpenAI Agents SDK and the Smol Agents framework for creating code agents.

Target Audience

The AI Engineering Bootcamp is designed for technically inclined professionals looking to delve into the practical application of AI:

• Software Engineers and Developers who want to learn how to build, deploy, and improve applications with LLM models in a production environment. The course will benefit individual engineers and team leaders to understand modern AI possibilities for their products.
• Data Specialists and ML Engineers (Data Scientists, Machine Learning/AI Specialists) aiming to master production development skills for applications based on machine learning and large language models. The program helps combine data analysis experience with engineering practices to create full-fledged AI systems.

To successfully pass the course, participants must have confident Python programming skills and an understanding of Data Science and prompt engineering basics. The training involves regular coding and active technical problem-solving - it might be challenging to recommend this intensive to programming novices or those unwilling to dedicate daily coding time.

Successfully completing the AI Engineering Bootcamp, a graduate achieves the following outcomes:

• Advancing as an AI Developer: You become a qualified AI Engineer - a specialist skilled in effectively applying modern AI tools in development. The course teaches you to be an AI-assisted developer who enhances their work with LLM and associated services. Mastery of advanced practices (the so-called lego blocks of creating AI applications) gives you the confidence to creatively solve tasks using AI at work or in personal projects.
• Ability to Create and Implement AI Applications: After Bootcamp, you will be able to independently build a fully-functional application based on large language models - from prototyping an idea to launching a service in production (in the cloud or on company servers).