Why evals in voice AI are so hard (and how to fix them)
Traditional metrics like WER often miss what matters in voice AI. Learn how to build custom evaluation frameworks that align with real user outcomes and product goals.
Raw WebSocket Voice Agent with AssemblyAIAssemblyai
Real-time STT latency benchmarks: what "fast enough" means for voice agentsAssemblyai
Analyzing and scoring voice agent calls with the LLM GatewayAssemblyai
Review - SimCLS and RefSum - Summarization TechniquesAssemblyai
Redact PII in Audio with Make and AssemblyAIAssemblyai
Review - Pretraining Representations for Data-Efficient Reinforcement LearningAssemblyai
Playing With P5.js: Creating a Voice-Controlled Game - Deepgram Blog ⚡️Deepgram
The Language of LGBTQ Inclusion and Allyship - Deepgram Blog ⚡️Deepgram
The Best (and Worst?) of Voice AI TechnologyDeepgram
No Code Transcription: Simplifying Workflows with Deepgram and Make.comDeepgram
Q&A with Deepgram’s New CPO, Ed AnuffDeepgram
How to Conduct an AI Agent Security Audit [+ Checklist]Zapier
Plumsail Documents: Zapier spp spotlightZapier
Oso Unbundles Security AuthorizationThenewstack
System UIconsCss Tricks
CSS Hell | CSS-TricksCss Tricks
Creating Color Themes With Custom Properties, HSL, and a Little calc()Css Tricks
pg_duckdb: Splicing Duck and Elephant DNAMotherduck
The rise of slow personal assistantsCerebras
Memo-ry: Simplifying Daily Tasks for People with Memory Loss - CerebrasCerebras
