Voice AI gateway — STT, LLM, and TTS through a single API, with real-time pipelines on LiveKit.https://speko.ai/https://speko.ai/blog/cutting-voice-agent-latency/https://speko.ai/blog/cutting-voice-agent-latency/A latency budget for cascaded voice pipelines, why endpointing is the silent killer, and where the architecture itself has to change when you cannot push lower.Thu, 23 Apr 2026 00:00:00 GMThttps://speko.ai/blog/designing-barge-in/https://speko.ai/blog/designing-barge-in/VAD is the load-bearing component. Most VADs are wrong for the job. A field guide to interruption that doesn't apologize, doesn't cough-trigger, and doesn't fall apart on a real phone call.Thu, 23 Apr 2026 00:00:00 GMThttps://speko.ai/blog/evaluating-voice-ai-production/https://speko.ai/blog/evaluating-voice-ai-production/Your benchmark says 5% WER. Your users say the agent can't understand them. Both are correct. The metrics that actually predict production failures.Thu, 23 Apr 2026 00:00:00 GMThttps://speko.ai/blog/framework-picking-stt-provider/https://speko.ai/blog/framework-picking-stt-provider/Six axes that decide whether your product ships — accuracy, latency, language coverage, cost, API ergonomics, vocabulary tolerance — with the tolerance thresholds we use to route traffic.Thu, 23 Apr 2026 00:00:00 GMThttps://speko.ai/blog/streaming-vs-batch-stt/https://speko.ai/blog/streaming-vs-batch-stt/Most of you shouldn't be using streaming STT. Four questions to answer honestly before you open another WebSocket.Thu, 23 Apr 2026 00:00:00 GMThttps://speko.ai/blog/voice-ai-noisy-audio/https://speko.ai/blog/voice-ai-noisy-audio/Noise is not one thing — it's four. A field guide to suppression, SNR thresholds, and the model choice that survives where your users actually live.Thu, 23 Apr 2026 00:00:00 GMT