Engineering notes — things I keep re-explaining, written down once.
- Single key, many models: what a switching layer really costs in a Node.js/Express API
- What I check before trusting async transcription webhooks for long support-call audio
- Node.js LLM summaries as a JSON contract: title, bullets, and action items
- Handling 429 rate limits and retries for speech-to-text APIs in Node.js
- Node.js text-to-image endpoint: prompt validation, signed URL or base64?
- A Node.js moderation pipeline that routes borderline user content to a review queue
- OpenAI-compatible API alternatives for an in-app chatbot: a one-key comparison
- OpenRouter vs direct OpenAI and Claude API: token cost, fallback and one key in Node.js
- Choosing a Speech-to-Text API for a SaaS App: REST, Pricing and Privacy