• What We Are Looking For

We want presentations that are practical yet impressive. Your proposal should promise a talk that gives attendees a "lightbulb moment" or a concrete blueprint they can apply to their own stacks the next day.

Optional Session Formats:

• 20/30 Minutes sessions on the main stage
• 50 Minutes Panel on the main Stage
• 5 Minute ignite talk about your AI Adoption gone horribly wrong (F*Ups talks) for the closing event.
• 1/2 Hour Workshops for the Code Labs track
• Deep Dive Case Study - 40 minutes + 15 Minutes of moderated Q&A from the audience. 

Suggested Topics:

Production-Grade LLM Ops: Dealing with latency, cost optimization, evaluation frameworks, and monitoring in the wild.

Advanced RAG & Vector Databases: Moving beyond naive RAG. Think chunking strategies, hybrid search, graph integration, and handling unstructured enterprise data.

Agentic Workflows that Work: Building, debugging, and steering multi-agent systems that actually complete complex business tasks without hallucinating.

Fine-Tuning & Local Models: When, why, and how to fine-tune smaller, open-source models (Llama, Mistral, etc.) to outperform generic frontier models.

The AI Edge: Deploying efficient models on-device, IoT, or constrained environments.

UX for AI: Brilliant, non-obvious interface design decisions that make AI features intuitive and reliable for human users.

Hard Lessons from the Trenches: "We tried X, it failed horribly, here is the architecture that actually fixed it."

Pro-Tip for Applicants: We judge proposals based on concrete details. Instead of "I will talk about how to use agents," try "How we built a 3-agent system using LangGraph that cut our customer support escalations by 40%, including code snippets of our state-management fix."

What We Are NOT Looking For

• Product demos and sales pitches 
• Soft skills