Deploy Python AI Automation System on Linux Server (Social + Multilingual Sub‑sites) Duration 3–4 working days Skills needed Python, Linux/Ubuntu, Nginx, MySQL, REST APIs, Social Media APIs (Meta Graph, X, Pinterest), S3‑compatible storage
About the project We run Librifly.com, a platform selling downloadable art, images, stories, poetry and maps, also available as posters, notebooks and books. We already have a complete AI automation system built: all Python code is written, tested and documented. This is not a coding project. You will deploy and configure the existing code on our Linux server and verify that the full system works end‑to‑end. Think of it as a technical installation and setup job: server environment, Nginx, database migrations, social media API configuration and final testing.
What the system does • Discovers content niches from Librifly.com (e.g. Monet, Vintage France travel, New York art). • Generates complete sub‑sites for each niche using AI (e.g. monet.librifly.com). • Populates each sub‑site with real products and pricing from Librifly.com. • Publishes each sub‑site in 9 languages, one new language per day. • Automatically posts to Instagram, Facebook, X (Twitter) and Pinterest when each site goes live.
Server environment • Hetzner Cloud server (Finland), Ubuntu + Nginx. • MySQL database already running. • Static IP and domain configured. • Python 3.12 and Redis to be installed (apt install).
Scope of work Step 1 — Server setup • Install Python 3.12 and Redis, create a virtualenv. • Install Python packages from requirements.txt. • Configure .env with our API keys. • Run 4 SQL migration files. • Set up Nginx wildcard subdomain (*.librifly.com). • Obtain wildcard SSL via Let’s Encrypt certbot. • Configure two systemd services (Celery worker and scheduler) to start on boot. Step 2 — Pipeline test • Run the pipeline manually and verify a sub‑site is generated. • Check HTML renders correctly in a browser. • Verify Nginx serves it on the correct subdomain. • Adjust scraper selectors if needed by inspecting Librifly.com. Step 3 — Social media setup • Create a Meta developer app (Instagram + Facebook). • Configure long‑lived access tokens for posting. • Set up X (Twitter) and Pinterest developer apps. • Configure a Hetzner Object Storage bucket (S3‑compatible). • Send at least one successful test post on each of the four platforms. Step 4 — Multilingual test • Run migration for language tables. • Build and test a Spanish version of one sub‑site. • Confirm hreflang SEO tags in HTML. • Check CJK (Chinese/Japanese/Korean) pages load with correct fonts.
Required skills • Python 3 (read/debug existing code). • Linux Ubuntu server admin. • Nginx (vhosts, wildcard subdomains). • SSL with Let’s Encrypt / Certbot. • MySQL (running SQL files, basic admin). • Social media APIs, especially Meta Graph API. • REST API basics (X and Pinterest). • S3‑compatible storage (e.g. Hetzner Object Storage).
What we provide • Full source code (4 documented zip files). • README files with step‑by‑step instructions. • SSH access to our server (securely shared). • All required API keys.
Deliverables 1. At least one sub‑site live at a .librifly.com subdomain. 2. Celery worker and scheduler running as system services and surviving restarts. 3. One successful test post on Instagram, Facebook, X and Pinterest. 4. Brief handover note covering token expiry dates and how to run the pipeline manually.
Meta setup note Meta (Instagram/Facebook) app setup is the most time‑consuming part because the portal is slow and the flow is non‑obvious. Please only apply if you have previously set up a Meta Graph API app; if not, say so clearly in your proposal.
How to apply Please include: 1. Whether you have set up a Meta Graph API app before (Instagram/Facebook posting). 2. Whether you have configured Nginx wildcard subdomains and Let’s Encrypt wildcard SSL. 3. When you can start and hours per day you can commit this week. 4. Your fixed‑price quote for the full project.
