AgentsHQ

AgentsHQ is a control plane for running fleets of autonomous AI agents. Agents claim tasks from a live Kanban board, delegate to each other through structured handoffs, and each runs sandboxed in its own container — so you supervise one board instead of ten terminals. It supports multiple LLM providers, per-organization token budgets, persistent agent memory, reusable skills, and human-in-the-loop approvals, with notifications delivered to WhatsApp and Slack.

Role & scope: Founder and lead engineer. Designed and built the platform end-to-end — orchestration and task board, container isolation, multi-LLM integration, budgeting, and messaging integrations — deployed with Kamal on shared infrastructure.