Slide 1

Slide 1 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration Christian Weyer | Co-Founder & CTO | Thinktecture AG | [email protected]

Slide 2

Slide 2 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration LLM- ALL-THE-THINGS? 2

Slide 3

Slide 3 text

Language Models understand and generate semantically rich human language, transforming it into text or structured data for both humans and machines. Embedding Models capture semantic meaning by encoding human language into numerical vector representations, facilitating understanding, comparison, and retrieval for both humans and machines. Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration 3 🫱 🫲 Semantic AI

Slide 4

Slide 4 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration MODELS FOR OUR SOFTWARE 4

Slide 5

Slide 5 text

§ Language models (LLMs, SLMs) are always part of end-to-end architectures § Everything goes into LM’s context via API calls § System prompt § User query § LMs enable human language as a first-class citizen 🤯 § Extending access possibilities to our software Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration End-to-end integration 5 Clients Services LLMs Desktop Web Mobile Service A Service B Service C API Gateway Monitoring LLM 1 LLM 2

Slide 6

Slide 6 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration Classical applications & UIs 6

Slide 7

Slide 7 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration EXTENDED END-TO-END SOLUTIONS 7

Slide 8

Slide 8 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration 8 C4 system context diagram

Slide 9

Slide 9 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration PATTERN RAG (RETRIEVAL-AUGMENTED GENERATION) 9

Slide 10

Slide 10 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration “Talk to your data” Cleanup & Split Text Embedding Question Text Embedding Save Query Relevant Results Question Answ er LLM 10 Embedding Model Embedding Model 💡 Indexing / Embedding Question Answering .md, .docx, .pdf etc. “Lorem ipsum…?” 💡 Vector DB

Slide 11

Slide 11 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration PATTERN STRUCTURED OUTPUT 11

Slide 12

Slide 12 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration Structured data from unstructured input - for API calling 12 “OK, when is my colleague CW available for a two- days workshop?” System Prompt (with employee data) + Schema / Function Calling (for structured output) (Web) API for Availability business logic

Slide 13

Slide 13 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration PATTERN SEMANTIC GUARDING & ROUTING 13

Slide 14

Slide 14 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration Semantics-based decisions for user interactions 14 Guarding (e.g. prompt injection) Routing (selecting correct target) “Lorem ipsum…?” Semantic Engine (Fine-tuned Language Model, Embedding Model) Target RAG 1 Target Structured Output & API Call Target RAG 2

Slide 15

Slide 15 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration PATTERN / SOLUTION OBSERVABILITY 15

Slide 16

Slide 16 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration Things can get… overwhelming 16

Slide 17

Slide 17 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration End-to-end tracing - with OpenTelemetry 17

Slide 18

Slide 18 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration END-TO-END SOLUTION ILLUSTRATED 18

Slide 19

Slide 19 text

Semantic routing Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration "Talk to your systems” 19 Web App / Watch App Speech-to-Text Internal Gateway (Python FastAPI) LLM / SLM Text-to-Speech Transcribe spoken text Transcribed text Check for experts availability with text Extract { experts, booking times } from text Structured JSON data (Function calling) Generate response with availability Response Response with experts availability 🔉 Speech-to-text for response Response audio Internal Business API (node.js – veeeery old) Query Availability API Availability When is CL…? CL will be…

Slide 20

Slide 20 text

Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration Recap: Top Semantic AI patterns & solutions – for end-to-end software engineering 20 RAG (Retrieval-Augmented Generation) Structured Output Semantic Guarding & Routing Observability

Slide 21

Slide 21 text

Thank you! Christian Weyer https://thinktecture.com/christian-weyer [email protected] 21

Slide 22

Slide 22 text

§ Technology catalyst § AI-powered solutions § Pragmatic end-to-end architectures § Microsoft Regional Director § Microsoft MVP for AI § Google GDE for Web AI [email protected] @christianweyer https://www.thinktecture.com Generative AI: LLM-basierte Anwendungen Top Patterns & Lösungen für nahtlose Integration Christian Weyer Co-Founder & CTO @ Thinktecture AG 22