shop.antonnovikov.com / llmops
AI observability and cost control
Once an AI feature is live, the same questions come up every week: what did it cost, why was that answer slow, and did the last prompt change make things worse. I add the monitoring I would put on any production service: every model call is recorded without personal data, cost is split by feature and by user, an alert fires as a budget runs low, and each prompt or model change is checked against a test set before it ships. Dashboards and alert rules come as files in your repository, so you can review and change them like any other code.
What is included
- Inventory of every place your code calls a model, with a feature name for each
- Every call recorded with model, tokens, cost, latency and outcome; personal data masked or never stored
- Dashboard of cost per feature and per user, with a projection to the end of the month
- Daily and monthly budgets with alerts to Telegram, Slack or e-mail
- A short note on the data: what is stored, for how long and who can see it
- Handover: source code and a short operations guide
Calculator
Request
To send a request, log in or create an account: you will see its status and can write to me in one thread. Your chosen options are kept.
Other services
- AI assistant for support and salesfrom $920 A chat assistant that answers customers from your knowledge, on your site or in messengers.
- Knowledge base search and Q&A (RAG)from $1,840 Ask questions in plain language, get answers with sources from your documents.
- AI workflow automationfrom $710 Routine work done by n8n or Python workflows with LLM steps: sort, summarise, draft, route.
- AI Telegram botfrom $1,050 A Telegram bot that chats, reads files and photos, remembers context and does things for users.