Marius Obert

Speech-to-Speech AI models - Marius Obert

Which speech AI is better: a modular system you control, or a 'magical black box' with lower latency? Learn the critical trade-offs for building real-time applications.

Speech-to-Speech AI models - Marius Obert
#1about 3 minutes

Comparing centralized transcription and direct speech-to-speech models

Two competing AI approaches are the centralized transcription model, which offers control but loses context, and the direct audio model, which is simpler but less transparent.

#2about 1 minute

The trade-off between developer control and model simplicity

Developers must choose between a transparent, customizable transcription pipeline and a "black box" speech-to-speech model that handles complexity automatically.

#3about 2 minutes

Understanding multimodal inputs and token costs in real-time AI

Real-time audio APIs can accept cheaper text tokens and even image streams as parallel inputs to provide richer context to the model.

#4about 3 minutes

Challenges in AI voice analysis and answering machine detection

While AI models can infer emotion from voice, they are not yet confident in their analysis, and reliably detecting answering machines remains a difficult problem.

#5about 1 minute

How to integrate speech-to-speech AI with the Twilio SDK

The Twilio SDK for the real-time API simplifies development by handling complex audio engineering tasks like codec conversion and sample rate management.

Related jobs
Jobs that call for the skills explored in this talk.

Featured Partners

Related Articles

View all articles
Chris Heilmann
Exploring AI: Opportunities and Risks for Developers
In today's rapidly evolving tech landscape, the integration of Artificial Intelligence (AI) in development presents both exciting opportunities and notable risks. This dynamic was the focus of a recent panel discussion featuring industry experts Kent...
Exploring AI: Opportunities and Risks for Developers
Chris Heilmann
With AIs wide open - WeAreDevelopers at All Things Open 2025
Last week our VP of Developer Relations, Chris Heilmann, flew to Raleigh, North Carolina to present at All Things Open . An excellent event he had spoken at a few times in the past and this being the “Lucky 13” edition, he didn’t hesitate to come and...
With AIs wide open - WeAreDevelopers at All Things Open 2025
Luis Minvielle
13 AI Tools for Developers
Artificial intelligence has rapidly transitioned from a hype item to a must-have tool for devs. Its adoption rate had seen a dramatic increase even before LLMs hit desktop computers, with AI in companies surging by 270% between 2015 and 2019.Develope...
13 AI Tools for Developers

From learning to earning

Jobs that call for the skills explored in this talk.