Systems Thinking in the AI Era, Course 1

Source: https://systemthinkinglab.ai/course-1

The course “System Thinking in the Age of AI — Course 1” is dedicated to the fundamental principles of designing software systems. It teaches a universal approach that allows for systematic design of nearly any digital platform — from social networks and streaming services to AI applications. The course is based on a framework of seven fundamental architectural building blocks, which has been taught for over ten years at UC Berkeley School of Information and is now available online.After completing the course, system design tasks stop being stressful improvisation. Instead of panicking at questions like “design Instagram,” you will be able to conduct a structured engineering conversation: clearly explaining architecture, arguing technological choices, analyzing trade-offs, and confidently managing AI tools when developing complex systems.The course is intended for technical specialists of different levels. Software engineers will learn to systematically respond to system design questions during interviews and stop guessing technologies blindly. Data Scientists and ML engineers will understand the infrastructure surrounding machine learning models: the model itself is only a small part of the system, and the main complexity lies in its scalable production. Bootcamp graduates will close a key gap between the ability to write a CRUD application and the ability to design a system intended for hundreds of thousands or millions of users.The main feature of the course is a universal architectural framework that is not tied to specific technologies. Unlike scattered materials on Redis, Kafka, or individual tools, the course offers a holistic language of system design: 7 building blocks from which any modern system can be assembled. These blocks represent enduring architectural patterns that have existed for decades and remain relevant regardless of changing technologies.Participants study three categories of components. Task processing blocks include services (APIs, web servers, and request/response interactions) and workers performing background tasks, long computations, and batch processing. Data storage blocks encompass key-value stores for fast queries and caching, file stores for media and CDNs, message queues for task management, relational databases for structured data and transactions, and vector databases for AI search and embedding operations. External entities that impact any system include users, external services (e.g., payments or third-party APIs), and time — a source of events, schedules, and automation.The course structure consists of 15 learning components. The fundamental part covers the basics of system thinking, the universal language of architectural blocks, their implementation in Python, and their mapping to real-world technologies. Practical labs demonstrate the operation of asynchronous systems through combinations of queues and workers, as well as automation through the interaction of time and background processes. A separate module is dedicated to external entities — users, third-party services, and temporal events — and their role in system architecture.The advanced part of the course teaches analysis of engineering trade-offs, methods for making architectural decisions, and recognizing patterns that emerge when combining building blocks. A practical case involves dissecting a URL shortening system, where participants conduct a full architectural analysis. The course concludes with a final evaluation of knowledge, including test and detailed questions.The final part of the program is a practical System Design Challenge, which takes place in three stages. First, participants design the minimal architecture of a social recipe platform. Then, they need to scale the system under conditions of viral audience growth. In the last stage, monetization features are added, requiring the evolution of the existing architecture and the adoption of new engineering decisions.As a result of completing the course, participants master a systematic approach to designing software platforms, learn to confidently discuss the architecture of complex systems, and acquire a fundamental language applicable to any modern product — from social networks to AI infrastructure.

Password: cms.ddpanda.org