Embarcadero Kai: Agentic Productivity and Delphi Development Modernization

Course taught by Daniele Teti; upon completion, a bit Time Academy certificate will be issued 

Duration: 1 day – Price: € 399.00 excluding VAT
Location: Online

Next upcoming course starts November 12, 2026

Course Description

The rise of agentic programming is redefining how we develop and maintain software in Delphi. This intensive one-day course demonstrates how to fully leverage Embarcadero Kai directly within the RAD Studio IDE to boost development team productivity, accelerate legacy code refactoring, and automate testing and bug-fixing phases.

Throughout the day, we will explore the entire AI-driven development lifecycle: from secure setup using local or BYOK cloud models, to building VCL/FMX desktop interfaces, reactive Full-Stack Web Apps (DMVCFramework + HTMX), and integrating external resources via the MCP (Model Context Protocol) standard.

Who Should Attend
  • Delphi Developers (Junior and Senior): Who want to integrate agentic development workflows and reduce coding and maintenance time by up to 40%.
  • Architects and Team Leads: Interested in evaluating the enterprise adoption of Kai while maintaining privacy and development ROI.
  • IT Companies: Looking to modernize VCL/FMX Delphi applications and speed up cost estimations and client proposals.
Prerequisites
  • Basic knowledge of Object Pascal / Delphi language.
  • General familiarity with the RAD Studio IDE and basic SQL or RESTful web development.
  • A subscription to an agentic coding service supported by KAI (e.g., Claude Code, Codex, Ollama Cloud, etc.).
Detailed Program
Morning Session: Setup, Agentic Workflow, and VCL/FMX Desktop UI (3.5 Hours)

 

Module 1: Kai Setup, Privacy, and “Agentic Mindset” (09:30 – 10:15)

  • RAD Studio Configuration: BYOK model (OpenAI, Claude, Gemini) and local LLMs via Ollama to protect company code and intellectual property.
  • Beyond the simple chatbot: Instructing Kai on project context using @file, @project, and advanced directives to let it operate autonomously in the IDE.
 

Module 2: Modern Desktop Development: VCL and FireMonkey (10:15 – 11:30)

  • VCL Interfaces: Rapid Form generation, data validation, and legacy code refactoring toward modern patterns (PPL, Async, Generics).
  • Multi-Device (FMX): Kai-guided adaptation and migration of code and UI from Windows to cross-platform (macOS, iOS, Android).
 

Coffee Break: 11:30 – 11:45

 

Module 3: The Agentic Loop: Build Error Fixing & Testing (11:45 – 13:00)

  • Autonomous Loop: Real-time automatic analysis of compilation errors with direct proposal and application of fixes in the source code.
  • Immediate Unit Testing: Automatic generation of unit tests (DUnitX) to safeguard the refactoring phase.
 

Lunch Break: 13:00 – 14:00

 
Afternoon Session: Web API, Full-Stack WebApp, MCP, and VCL Workshop (3.5 Hours)

 

Module 4: Web Development with Kai: Web API and WebApp (14:00 – 15:15)

  • Web API with DMVCFramework: Kai-driven generation of RESTful Controllers, DTOs, and routing starting from informal specifications.
  • Full-Stack WebApp (DMVCFramework + HTMX): Rapid development of reactive web interfaces by delegating the writing of server-rendered HTML/HTMX templates.
 

Module 5: Advanced Data Connection with MCP (15:15 – 16:00)

  • Beyond the IDE: Using the MCP (Model Context Protocol) standard to connect the AI agent to external resources and enterprise databases.
  • Practical Example: Automatic analysis of an external DB schema via MCP with on-the-fly generation of DataModules and persistence logic.
 

Short Break: 16:00 – 16:15

 

Module 6: Hands-On Workshop: End-to-End Exercise on VCL App (16:15 – 17:00)

  • VCL Practical Project: Modernization or creation from scratch of a real-world VCL application module.
  • Complete Workflow: Data schema reading, VCL UI generation, asynchronous business logic refactoring, and automated bug resolution via the agentic loop.
  •  

Module 7: Enterprise ROI, Cost Estimation, and Q&A (17:00 – 17:30)

  • Business Analysis: How agentic AI reduces VCL development and maintenance time by up to 40%.
  • Strategic Evaluation: How to leverage AI speed in development estimates and final client quotes.
  • Final Q&A and bit Time Academy Certificate Awarding.
Included Course Materials
  • Course slides and in-depth documentation.
  • Repository containing all source code from examples and exercises.
  • Official bit Time Academy Certificate of Completion.
RESERVE YOUR SEAT
  • Completing the form does not guarantee a seat for the upcoming course. The course fee is € 399.00 excluding VAT.
Scroll to Top