https://anovikau.com/writing/ Posts on distributed systems, real-time advertising, reliability and side projects. en Sun, 25 Oct 2026 00:00:00 GMT https://anovikau.com/writing/tape-archive/ https://anovikau.com/writing/tape-archive/ Sun, 25 Oct 2026 00:00:00 GMT Six hours of family tapes from 1980 to 2000, catalogued on a CPU: songs, talk, dates said out loud, and one song found on five different tapes. What real recordings taught me about thresholds and hallucinations. Experiments Python Whisper Audio Open source Side project https://anovikau.com/writing/users-moved-to-sweden-overnight/ https://anovikau.com/writing/users-moved-to-sweden-overnight/ Thu, 22 Oct 2026 00:00:00 GMT Analytics said a big share of Paperspell's users suddenly lived in small Swedish towns. They weren't bots and they weren't Swedish: a floating SDK version had started sending events from 127.0.0.1. Engineering Product analytics PostHog Flutter Mobile Debugging https://anovikau.com/writing/cookieless-is-not-anonymous/ https://anovikau.com/writing/cookieless-is-not-anonymous/ Mon, 19 Oct 2026 00:00:00 GMT Chrome kept third-party cookies, yet ads still recognise you. How Unified ID 2.0 turns the email you log in with into a token, what genuinely improved, and what didn't. Engineering Privacy AdTech Cookieless UID2 Data protection https://anovikau.com/ru/writing/kola/ https://anovikau.com/ru/writing/kola/ Sun, 18 Oct 2026 00:00:00 GMT Как программист из Варшавы собрал мужской круг у костра на Висле: откуда взялась идея, чему я учился как фасилитатор, как прошли первые две встречи и почему я в этом круге не учитель, а участник. Life Кола Эмиграция Группы равных Варшава https://anovikau.com/writing/artf-agents-in-the-auction/ https://anovikau.com/writing/artf-agents-in-the-auction/ Fri, 16 Oct 2026 00:00:00 GMT IAB Tech Lab's Agentic Real-Time Framework turns the ad auction into a platform for agents from any vendor. Four places where its contract still hurts, and why MCP belongs on the orchestrator. Engineering AdTech RTB AI agents MCP System design https://anovikau.com/writing/my-ai-memory/ https://anovikau.com/writing/my-ai-memory/ Wed, 14 Oct 2026 00:00:00 GMT 5,362 ChatGPT conversations that Claude knows nothing about. An experiment in a memory that belongs to me rather than to an AI provider: plain Markdown in a private git repository, readable by any assistant. The design, and what a 5% pilot showed about cost and quality. Experiments AI Personal AI Knowledge management LLM https://anovikau.com/writing/the-nda-was-in-a-git-branch/ https://anovikau.com/writing/the-nda-was-in-a-git-branch/ Thu, 08 Oct 2026 00:00:00 GMT A friendly recruiter, a real-looking healthcare project and an NDA that could only be read with git checkout. Inside the tarball: a post-checkout hook that would have run a stranger's script on my laptop. What the hook does, which git commands trigger it, and how I open such archives now. Experiments Security Git AI agents Hiring scams https://anovikau.com/writing/anthropic-interviewer/ https://anovikau.com/writing/anthropic-interviewer/ Wed, 07 Oct 2026 00:00:00 GMT Anthropic Interviewer, Claude asking questions for Anthropic's research team, asked me how AI is changing my work and my life. The full conversation, translated from Russian: why running several agents at once feels almost addictive, what I don't want AI to change, and who should share in what it creates. Life AI Anthropic Work Society https://anovikau.com/ru/writing/anthropic-interviewer/ https://anovikau.com/ru/writing/anthropic-interviewer/ Wed, 07 Oct 2026 00:00:00 GMT Anthropic Interviewer, Claude, который задаёт вопросы от имени исследовательской команды Anthropic, расспросил меня, как ИИ меняет мою работу и жизнь. Полный текст разговора: почему от агентов так трудно оторваться, что я не хотел бы, чтобы ИИ менял, и кому должны достаться его блага. Life ИИ Anthropic Работа Общество https://anovikau.com/writing/jev-on-my-inbox/ https://anovikau.com/writing/jev-on-my-inbox/ Fri, 02 Oct 2026 00:00:00 GMT Jev, the new model from TypeSafe AI, answers typed questions with calibrated probabilities instead of text. On 104 of my real emails it was four times faster than Gemini Flash-Lite and nearly as accurate, and as a pre-filter it would have skipped 35–42% of LLM calls. Why my email agent still doesn't use it. Experiments AI LLM Open source Privacy Email agent