Real project examples: what was done, with which tools and with what result.
A company with hundreds of internal documents. I deployed a private RAG: documents are split into fragments, searched by meaning, and a local model writes the answers. Nothing leaves the server.
Search < 2 s · 100% privateA live project moved to a new server with zero downtime: Docker containers, Nginx proxy, fail2ban, nightly backups and monitoring. No incidents since.
0 downtime · nightly backupsAn LLM-powered bot answers typical questions about services and prices, captures leads and hands complex cases to a human. It closes up to 80% of requests.
-80% support routineBuilt a monitoring stack: Grafana, server and service metrics, Telegram alerts with context. I now spot downtime before clients do.
Reaction time: minutesInvoices and contracts are generated automatically from CRM data: template + substitution + delivery to the client. It used to take half a day.
Hours of routine → 2 minutesDeployed Ollama with Open WebUI on a private server: private chats, text generation, work with documents. API costs cut up to 10 times.
Up to 90% saved on APIsTell me what hurts - I will suggest a solution and an honest estimate. The first consultation is free.
Discuss your project