IDEAS, DECISIONS & LESSONS
Notes on AI, architecture and building useful products.
32 articles
01 Mar 2026 · 15 min read
ai
Part 11 of 12 — The real cost of AI voice interviews, broken down per minute. Managed vs self-hosted economics, the three tipping points, and how to get from $3.45 per interview to under $1.00.
Read more28 Feb 2026 · 18 min read
architecture
Part 10 of 12 — From 10 concurrent interviews to 10,000. LiveKit SFU mesh, stateless agent workers, Kubernetes auto-scaling, regional deployment, and the infrastructure patterns that handle hiring season surges.
Read more28 Feb 2026 · 12 min read
agents
Part 3 of 6 — Research interviews follow structured protocols with distinct phases. How to build an LLM-driven state machine with next_phase() function calling and dynamic instruction swapping via set_chat_ctx().
Read more27 Feb 2026 · 5 min read
azure
Part 1 of 7 — Why Azure Foundry Voice Live changes everything for real-time voice AI. Understand the architecture, compare alternatives, and see how to build an interview system that responds in under 200ms.
Read more27 Feb 2026 · 5 min read
audio-quality
Part 5 of 7 — Low latency is useless if the voice sounds robotic. Learn how to configure audio codecs, enable noise suppression, handle natural interruptions, and choose the right AI voice persona for an interview system.
Read more27 Feb 2026 · 6 min read
azure
Part 6 of 7 — Things will break. This is your complete troubleshooting guide for Azure Voice Live in Next.js: WebSocket drops, choppy audio, VAD false positives, session timeouts, quota limits, and every other issue you'll encounter.
Read more27 Feb 2026 · 6 min read
azure
Part 7 of 7 — Your demo works. Now make it production-ready for 1,000 concurrent interviews. Full guide to deployment options, WebSocket scaling, Azure pricing breakdown, cost estimations, and monitoring for Azure Voice Live.
Read more27 Feb 2026 · 16 min read
ai
Part 9 of 12 — Recording interviews is easy. Doing it legally isn't. LiveKit Egress, real-time transcription, GDPR consent flows, HIPAA encryption requirements, and the compliance checklist you need before go-live.
Read more27 Feb 2026 · 5 min read
azure
Part 4 of 7 — Every millisecond counts in a voice interview. Learn how to architect your Azure Voice Live system for sub-150ms end-to-end latency through region selection, chunk tuning, buffer strategies, and WebSocket optimization.
Read more
