Builders Workshop: Real-time AI Voice Agents

​Step into the AWS GenAI Loft for a deep‑dive, 200/300‑level workshop where you’ll learn how to architect and build a live, real‑time AI voice agent from the ground up. Our three expert presenters will guide you through every piece of the puzzle—speech in, AI brains, and speech out—so you can walk away with working code and best practices for production.

What You’ll Learn

• ​Ultra‑Low‑Latency Speech I/O: Integrate Deepgram’s STT and TTS for sub‑second transcription and natural‑sounding synthesis.
  
  

• ​Pipeline Orchestration with Pipecat: Use Pipecat’s open‑source framework to chain audio, vision, and LLM steps into a seamless conversational flow.
  
  

• ​Advanced LLM Logic with Nova Pro: Leverage Amazon’s newest Nova Pro model to manage context, handle branching logic, and generate dynamic responses.
  
  

• ​Deployment & Monitoring Tips: Learn best practices for scaling your voicebot, instrumenting telemetry, and ensuring reliability in production.
  

​Who Should Attend

​This workshop is perfect for:

• ​Developers & Engineers building voice‑driven applications
  
  

• ​AI/ML Practitioners looking to operationalize speech and language models
  
  

• ​Solution Architects designing real‑time, multimodal systems
  
  

• ​Tech Leads & Innovators exploring GenAI‑powered user experiences

​Agenda

• ​5:00 PM- 5:30 PM   Welcoming
  
  

• ​5:30-6:30 Hands on keyboard walk through of building real time AI voice agents
  
  

• ​6:30 PM- 7:00 PM   Live Q&A, networking and food provided

​Speakers

• ​Kwin Hultman Kramer
  CEO & Founder, Daily & Pipecat
  

• ​Natalie Rutgers
  Vice President of Product, Deepgram
  

• ​Ayan Ray and Eric Yuen
  GenAI Specialists, AWS
  
  

​This event is brought to you by Daily and Deepgram.