I'm the founder of Scalably. I build and run multi-tenant AI agent platforms inside the operations of real companies across Southeast Europe. The agents work through Slack, Telegram, WhatsApp, and the web, and they ship finished work rather than stop at answers.
My focus is the unglamorous part: per-tenant isolation, MCP tool integration, model routing, persistent memory, approval boundaries, delivery verification, and the failure modes that appear only when agents touch real business systems. Scalably's current production runtime uses the Claude Agent SDK. I build with Claude Code and Codex: coding, orchestration, research, and independent verification of each other's work.
What I work on
Scalably's current production runtime is built on the Claude Agent SDK. It runs across multiple channels (Slack, Telegram, WhatsApp, web), exposes a governed catalog of MCP servers and permissioned skills to each tenant, and routes between Claude models by measured task performance rather than preference. I author MCP servers as production artifacts and use Codex to build, inspect, and verify work across the stack.