Summary
Stephanie Nyarko walks through the end-to-end process of deploying the Hermes AI agent — a self-improving agentic assistant developed by Nous Research — on AWS Lightsail, creating a personal AI assistant that can run scheduled research tasks like competitor YouTube channel scans and content performance summaries before you start your workday.
The tutorial begins with AWS Lightsail instance creation, where Nyarko recommends selecting a region close to your physical location for latency reasons and choosing at least the 2GB memory plan to avoid agent crashes — a common mistake she flags from community research. She opts for Ubuntu on a general-purpose instance and connects via the browser-based SSH interface, avoiding the need for key files or local terminal apps. Hermes is installed via a single Linux command, with setup taking 10–15 minutes autonomously.
From there, the video covers the Quick Setup path, which is free and requires no API keys, as well as how to select from free models or paid providers like Anthropic and OpenAI once usage demands grow. Tool selection is also explained — web search for research tasks, text-to-speech for audio output — and the local terminal backend is recommended for standard single-server deployments. The result is a cloud-hosted AI assistant accessible over Telegram that runs 24/7 without occupying a local machine. The tutorial is accessible to non-technical creators and small-business owners who want autonomous AI workflows without managing complex infrastructure.
📺 Source: Stephanie Nyarko · Published July 13, 2026
🏷️ Format: Tutorial Demo
