Description
Summary We are looking for an AI agent engineer to extend an already operational Hermes agent with executive-assistant and automation capabilities. This is not a new agent built from scratch: all requested features are extensions to the current Hermes runtime and its existing Telegram-based workflow. The priority scope includes smarter prompt enrichment and model routing, isolated long-term memory, a lightweight admin dashboard, secure credential handling, controlled browser automation, and selected communication, scheduling, research, or specialist RAG integrations. External messages, public posts, purchases, account changes, and other sensitive actions must use explicit approval gates. See details in the attached file. Expected delivery is 5–10 working days, depending on the condition of the current codebase and availability of API access and test accounts. The contractor should audit the existing setup on Day 1, implement the highest-value features, test failure and security paths, and provide deployment notes plus a clear Phase 2 backlog. Please include relevant AI-agent projects, your Hermes experience (it's mandatory), your proposed 5-day versus 10-day prioritization, availability, and estimate.