Talk | 
EKON 28
 | 04.11.2024

Real-World RAG mit eigenen Daten & Dokumenten – Generative AI, Semantische Suche & LLMs jenseits von “Hallo Welt!”

Semantische Suche ist wohl das Paradebeispiel, um Generative AI mit eigenen Daten in eigene Businessanwendungen einzubinden. Mindestens eine Fantastilliarde Blogposts, YouTube-Tutorials und sonstige Artikel zeigen, wie man so etwas mit Large Language Models (LLMs) und Retrieval-Augmented Generation (RAG) - offenbar sehr einfach - realisieren kann. Doch wie immer steckt der Teufel im Detail. In dieser Session führt Sie unser Experte Sebastian Gingter durch ein bisschen Theorie zu Vektoren und Embeddings, zeigt, was Vektordatenbanken leisten und was Splitting beim Erzeugen von Embeddings bedeutet. Worauf müssen wir beim Indizieren achten? Dabei buhlen noch die verschiedensten Tools wie LangChain, Azure Cognitive Search, Qdrant oder Chroma um unsere Aufmerksamkeit. Sehen Sie, wie man am Ende alle Schritte in der eigenen Applikation zu einem effizienten RAG-Prozess zusammenführt, der dann auch wirklich sinnvolle Antworten auf die Fragen der Benutzer zu den Daten & Dokumenten Ihrer Anwendung liefert.

Sebastian ist Consultant bei Thinktecture und Microsoft MVP. Seine Spezialgebiete sind Generative AI im Businessumfeld sowie ASP.NET Core.

Event

EKON 28
04.11.24  
- 06.11.24 
@ Düsseldorf
 (DE)
EKON_187x50

Slidedeck

Weitere Artikel zu aktuellen Schwerpunkten

.NET
pg

Discriminated Unions in .NET: Modeling States and Variants

Domain models often involve concepts that exist in multiple distinct states or variations. Traditional approaches using enums and nullable properties can lead to invalid states and scattered logic. This article explores how discriminated unions provide a structured, type-safe way to model domain variants in .NET, aligning perfectly with Domain-Driven Design principles while enforcing invariants at the type level.
06.10.2025
.NET
pg

Discriminated Unions in .NET: Modeling States and Variants

Domain models often involve concepts that exist in multiple distinct states or variations. Traditional approaches using enums and nullable properties can lead to invalid states and scattered logic. This article explores how discriminated unions provide a structured, type-safe way to model domain variants in .NET, aligning perfectly with Domain-Driven Design principles while enforcing invariants at the type level.
06.10.2025
.NET
pg

Smart Enums in .NET: Integration with Frameworks and Libraries

Learn how to seamlessly integrate Smart Enums with essential .NET frameworks and libraries. This article covers practical solutions for JSON serialization, ASP.NET Core model binding for both Minimal APIs and MVC controllers, and Entity Framework Core persistence using value converters. Discover how Thinktecture.Runtime.Extensions provides dedicated packages to eliminate integration friction and maintain type safety across your application stack.
21.09.2025
.NET
pg

Value Objects in .NET: Enhancing Business Semantics

Value objects are fundamental building blocks in Domain-Driven Design, serving far more than simple data wrappers. This article explores their strategic importance in bridging technical code and business concepts, enforcing domain rules, and fostering clearer communication with domain experts. Learn how to build robust aggregates, cultivate ubiquitous language, and encapsulate domain-specific behavior using Thinktecture.Runtime.Extensions in .NET applications.
16.09.2025
.NET
pg

Pattern Matching with Discriminated Unions in .NET

Traditional C# pattern matching with switch statements and if/else chains is error-prone and doesn't guarantee exhaustive handling of all cases. When you add new types or states, it's easy to miss updating conditional logic, leading to runtime bugs. The library Thinktecture.Runtime.Extensions solves this with built-in Switch and Map methods for discriminated unions that enforce compile-time exhaustiveness checking.
26.08.2025
.NET
pg

Value Objects in .NET: Integration with Frameworks and Libraries

Value Objects in .NET provide a structured way to improve consistency and maintainability in domain modeling. This article examines their integration with popular frameworks and libraries, highlighting best practices for seamless implementation. From working with Entity Framework to leveraging their advantages in ASP.NET, we explore how Value Objects can be effectively incorporated into various architectures. By understanding their role in framework integration, developers can optimize data handling and enhance code clarity without unnecessary complexity.
12.08.2025

Unsere Webinare

Unsere Artikel

Mehr über uns