Modern AI solutions become dramatically more valuable when enriched with context—local knowledge and insights that AI models can't inherently possess from training alone. This talk will explore various practical methods for injecting essential local context into AI, enabling better results, more specialized for the specific scenario.

Discover how techniques such as Retrieval-Augmented Generation (RAG), embeddings, vector databases, fine-tuning, and targeted tool-calling significantly increase the precision and determinism of AI outputs. Learn how sophisticated agentic behaviors and tool calls, including integrations like MCP, turn passive AI interactions into proactive, genuinely useful experiences.

We'll address the challenges and solutions involved in optimizing AI solutions across multiple models, highlighting both immediate strategies and long-term approaches that help getting the desired results from the AI. Join us to unlock the full potential of context-driven AI and create smarter, more assistive experiences.