Build with AI Wrocław: The Full-Cycle Journey

Location: Uniwersytet Ekonomiczny we Wrocławiu

Dates: June 29th – July 4th, 2026

The era of “that’s not my job” is over. Welcome to Build with AI Wrocław, an intensive, 6-day marathon designed to evolve UI/UX Designers, Developers, and DevOps Engineers into cross-functional Product Creators, leveraging AI as a collaborative bridge across the entire SDLC.

DAY 1 - 29.06 · Product Design, Scope & MVP

Move from a blank page to a scoped MVP you'll keep building all week:

• Define a real user persona and a real problem statement.

• Map the core flow of your product and determine the scope for your MVP.

• Slice it down to what's actually shippable, learning where AI helps and hurts.

Schedule:

• 4:00 PM · The use of AI should not be a goal in itself. When AI truly creates value for users and businesses, and when it leads to products that fail to solve any meaningful problem.

• 4:55 PM · Break

• 5:00 PM · From idea to product – understanding the IT project lifecycle. Introduction to Design Thinking and the SDLC: why the order of steps matters and what happens when teams skip Discovery.

• 5:15 PM · Who is your user? Building a persona as the foundation for every product decision, avoiding features that nobody needs.

• 5:30 PM · Break

• 5:35 PM · Workshop exercise: Problem Statement Canvas + Persona. Teams define their target user and formulate a one-sentence problem statement for validation.

• 6:35 PM · BPMN for the impatient – how to visualize a business process before designing a solution.

• 6:50 PM · Event Storming.

• 7:05 PM · Break

• 7:20 PM · Modeling your product’s process using the selected method (BPMN or Event Storming). Teams map the workflow of their solution.

• 8:10 PM · Time to create tasks – building a Kanban board. Translating developed artifacts (persona, problem statement, process diagram) into concrete tickets ready for implementation.

• 8:40 PM · Q&A Session.

• 9:00 PM · Workshop Closing.

Trainers: Dawid Perdek, Natalia Kierczuk, Artur Suchwalko

DAY 2 - 30.06 · UI/UX & AI-Native Design Systems

Turn product thinking into tangible, inclusive user interfaces:

• Learn the fundamentals of digital accessibility and why inclusive design matters.

• Understand how AI helps create, scale, and maintain a reusable design system.

• Connect design decisions directly to implementation through AI-powered workflows and MCP servers.

Schedule:

• 4:00 PM · What is accessibility and why does it matter? Introduction to digital accessibility, its role in creating inclusive digital products, and how to incorporate it from the very beginning.

• 4:55 PM · Break

• 5:00 PM · AI-Native Design Systems: Building Products With and Without a Designer. How developers can generate a functional design system from scratch using Antigravity and bring it into Figma through AI-powered workflows. Using AI to understand design tokens, styles, and component anatomy to keep generated interfaces consistent with original design intent.

• 7:05 PM · Break

• 7:20 PM · From Design System to Working Product. Transforming your design system into the first working version of a product. Exploring how a coding agent treats the design system as a source of truth via MCP servers, automating design decision inheritance in generated code.

• 8:40 PM · Q&A Session.

• 9:00 PM · Workshop Closing.

Trainers: Kinga Witko, Witold Iglewski, Przemyslaw Nowak

DAY 3 - 01.07 · Scaling Frontend Architecture & AI-Assisted Debugging

Scale application complexity under control using robust architectures and advanced diagnostic workflows:

• Master Nx monorepo boundaries to isolate and share enterprise frontend code.

• Harness modern Angular reactivity and signal-driven design to manage clean state.

• Build performance-driven, interactive node-and-edge diagrams using native signals in ng-diagram.

• Supercharge debugging with AI agents connected directly to Chrome DevTools and source context.

Schedule:

• 16:00 · Enterprise Modularization: Introduction to Nx. Understanding workspaces, dependency graphs, and how to structure a frontend repository that scales without turning into a monolith.

  Trainer: Julia Rapczynska

• 17:00 · Modern Angular Foundations. Diving into modern Angular architecture, component design, and signal-based reactive patterns to build high-performance user interfaces.

  Trainer: Adrian Romanski

• 17:50 · Break (10 minutes)

• 18:00 · Native Diagramming with ng-diagram. Building interactive canvas interfaces, workflows, and custom node charts. Exploring ng-diagram's open-source, HTML/CSS-first architecture and its signal-based state to create responsive visual data applications natively in Angular.

  Trainer: Jakub Kubacki

• 19:00 · Break (15 minutes)

• 19:15 · Application Built - What Next? Debugging and Optimization in the AI Era. Diagnosing layout, accessibility, and rendering problems efficiently using Antigravity, Chrome DevTools MCP, and Modern Web Guidance to give AI agents (and yourself) the context needed to instantly fix bugs.

  Trainer: Dominika Zając

• 21:00 · Workshop Closing.

DAY 4 - 02.07 · Fullstack, Architecture & LLM Engineering

Bridge the gap between frontend, backend, and non-deterministic AI execution:

• Connect an Angular frontend to a custom NestJS backend inside an Nx monorepo.

• Spin up persistent database storage using Docker and Sequelize ORM.

• Break the single-prompt ceiling by implementing evaluation scoreboards.

• Build autonomous agentic loops using custom code workflows and tool calls.

Schedule:

• 16:00 – 16:05 · NestJS and API Fundamentals (5 min) · Basics of NestJS, why it resembles Angular, and core definitions (requests, HTTP methods, statuses, CORS).

• 16:05 – 16:15 · Monorepos with Nest and Angular (10 min) · Reviewing Nx, tags, and context boundaries to maintain architecture order as the project grows.

• 16:15 – 16:20 · How Angular Communicates with the Backend (5 min) · Navigating App Initializer, Http Interceptor, and Http Client.

• 16:20 – 16:30 · Lab 1: Launching the Project (10 min) · Running the app locally and checking the boilerplate monorepo setup running with a json-server fake backend.

• 16:30 – 16:50 · Break (20 min)

• 16:50 – 17:05 · NestJS Architecture Deep Dive + OpenAPI (15 min) · Building blocks (modules, controllers, providers, middleware, guards) and configuring automated OpenAPI/Swagger documentation.

• 17:05 – 17:25 · Lab 2: Your First Real Backend (20 min) · Creating a custom users module, writing endpoints in NestJS, and repointing the frontend away from the mock server.

• 17:25 – 17:30 · Break (5 min)

• 17:30 – 17:45 · Baza danych i ORM: od obiektów do rekordów (15 min) · Connecting code objects to functional SQL database tables via Sequelize.

• 17:45 – 18:00 · Lab 3: From Mocks to Databases (15 min) · Spinning up a database with Docker, configuring Sequelize, and checking the full Frontend $\rightarrow$ Backend $\rightarrow$ Database pipeline flow.

• 18:00 – 18:30 · Evals 101: How to Evaluate a Machine That Improvises (30 min) · Mapping the four quadrants of evaluation, separating tests from evals, and running a live scoreboard.

• 18:30 – 19:00 · Prompt Engineering: Teaching the Model the Rules of the Game (30 min) · Coding rules directly into prompts to increase scores, tracking prompt bloat, and finding model breaking points.

• 19:00 – 19:30 · Context Engineering: Stop Sending the Entire Instruction (30 min) · Selecting and passing only task-relevant tokens instead of packing entire rulebooks into the context window.

• 19:30 – 19:45 · Break (15 min)

• 19:45 – 20:05 · Harness Engineering I: One Agent, Real Tools (20 min) · Moving away from single-shot calls to an iterative agent loop that can execute a plan $\rightarrow$ build $\rightarrow$ verify $\rightarrow$ fix workflow.

• 20:05 – 20:45 · Harness Engineering II: Multi-Agent Systems in Code (40 min) · Writing code-controlled workflows (loops/conditions) to split tasks between isolated, single-responsibility agents.

• 20:45 – 21:00 · What's Next: Tools Beyond the Workshop (15 min) · Mapping out enterprise-grade tools for production codebases.

Trainers: Greg Radzio, Adam Kowalski

DAY 5 - 03.07 · Deployment, Scale & Optimization

Expand your MVP's capabilities with specialized tools and deploy securely to production:

• Build a custom MCP server to give LLMs access to advanced external tools.

• Integrate pytriz to give models the ability to resolve technical contradictions.

• Host your full stack (Nx + Angular + NestJS + SQL) and MCP server on Google Cloud.

• Analyze system health and optimize performance with live cloud logs.

Schedule:

• 16:00 · Building Advanced AI Tools: Creating an MCP Server for Technical Problem Solving.

  Hands-on deep dive into Model Context Protocol (MCP). Using a template boilerplate repository, Docker configs, and the pytriz Python package, you will build, layout, and run a dedicated MCP server to expose capabilities to an LLM designed to handle technical contradictions using TRIZ.

  Trainer: Marek Mysior

• 17:30 · Break (15 minutes)

• 17:45 · Enterprise Cloud Deployment & Cloud Native Architectures.

  Transitioning your full development stack into production using Google Cloud best practices.

  • Automated CI/CD: Packaging secure container images with Cloud Build and storing them in Artifact Registry.

  • Serverless Scale: Deploying the frontend, backend, and your new TRIZ MCP server natively on Google Cloud Run with explicit environment configs and IAM permissions.

  • Cloud Data: Connecting the NestJS backend securely to a managed Cloud SQL instance.

  • Observability: Setting up dashboards, configuring alerts, and checking Cloud Logging to triage runtime bugs under heavy load.

    Trainers: Moksh Atukuri, Jacek Kubiak

• 20:20 · Break (10 minutes)

• 20:30 · Hackathon Task Presentation & Ideation.

  Introduction to the hackathon challenges. Teams analyze technical guardrails, pick their target problem, and architect how they will link their full system components (Nx, Angular, NestJS, SQL, and MCP tools) into a working prototype.

• 21:00 · Workshop Closing.

DAY 6 - 04.07 · The Grand Hackathon

The ultimate test of collective AI superpowers. Cross-functional teams collaborate intensely to design, build, and completely launch a real product.