Avatar

Promptly Done

Explorations in Voice AI.

    1. Dark Mode
Voice Agents

Cutting LLM Latency in Half with Voice Agents

How to cut your LLM's TTFT latency by 60%

Aug 13, 2025
7 minute read
Voice Agents

Building a Low Latency (Under 1000 ms) Arabic Voice Agent with RAG Using Ultravox & LiveKit

How to build ultra-low latency Arabic voice agents using end-to-end speech models like Ultravox, with tool use and RAG capabilities

Jul 29, 2025
5 minute read
Audio ML

Experimenting with Speech Augmentations: Enhancing Speech-to-text Model Robustness

Explore speech augmentation experiments to boost ASR model robustness. Learn key techniques, practical examples, and their impact on WER/CER

May 13, 2025
8 minute read
Audio ML

Curating Custom Datasets for Arabic Speech-to-text Models:

A Case Study on what not to do: Lessons learned curating diverse Egyptian Arabic speech datasets for training high-quality ASR models

Apr 28, 2025
13 minute read
© 2025 Promptly Done
Built with Hugo
Theme Stack designed by Jimmy