Ever wondered what powers GripAi behind the scenes? We built a short promo video that walks through the entire ecosystem — every service, every data pipeline, every automated job that keeps the platform running 24/7.
Watch the Video
This 105-second narrated walkthrough covers every layer of the GripAi infrastructure — from the three public domains to the backend services processing thousands of data points daily.
The Three Domains
The GripAi ecosystem runs across three interconnected domains, each serving a different purpose:
- GripAi.uk — The business face. The commercial landing page and partnership hub.
- jaffaAi.cc — The brain. Our AI-powered incident intelligence engine that analyses gaming issues in real time.
- GameGrip.cloud — The front. The blog you’re reading right now, plus the sandbox environment and public API.
20 Backend Services — All Online, All the Time
Under the hood, GripAi runs 20 PM2-managed Node.js services on an Ubuntu 24.04 server. These include:
- jaffa-ai — The core AI intelligence engine
- gamegrip-api — Dual-lane API serving both internal and public requests
- risk-api & risk-worker — Real-time PSI risk scoring for game issues
- recommend-api & recommend-worker — AI-powered fix recommendations
- blog-api & blog-worker — Automated blog content pipeline
- driver-api — GPU driver intelligence tracking
- website-api — Website monitoring across 50+ gaming sites
- sorter-api & sorter-worker — Data classification and ranking
- temporal-api & temporal-worker — Time-series trend analysis
- …and 7 more specialised services
Total memory footprint: just 637 MB. Efficient by design.
The Data Pipeline: Crawl → Process → Publish
Seven automated crawlers pull data from across the gaming internet:
- YouTube gaming channels
- Reddit gaming threads
- Steam forums and community hubs
- GitHub issue trackers
- Hardware and driver release feeds
- Mobile gaming feeds
- 50+ gaming websites monitored every 5 minutes
Five processing engines then sort, score, cluster, and correlate this data before publishing to six output channels: this blog, Bluesky, Mastodon, Discord, the jaffaAi dashboard, and email newsletters.
The Database: 620,000+ Rows of Intelligence
All this data feeds into a MySQL database with 41 tables containing over 620,000 rows of gaming intelligence. The biggest tables include:
- cluster_trends — 207,000+ trend data points
- cluster_history — 188,000+ historical records
- cluster_recommendations — 50,000+ AI-generated fix suggestions
- evidence — 38,000+ pieces of supporting evidence
- blog_posts — 6,897 auto-generated articles
41 Cron Jobs — Zero Manual Intervention
The entire system is automated with 41 cron jobs running around the clock. Content crawlers fire every 4-8 hours. Hardware and driver scans every 6 hours. Risk scoring, blog generation, and social media posting all run on their own schedules. The system never sleeps.
One Ecosystem. Always On.
5 domains. 33+ pages. 20 services. 41 cron jobs. 41 database tables. 620,000+ data rows. All connected. All automated. All running 24/7.
That’s the machine behind the magic.
Follow GripAi
🦋 Bluesky — @gripaiuk.bsky.social
🐘 Mastodon — @gripai@mastodon.social