CI demand surges 9x in a week as AI-generated code floods pipelines, Celesto shares scaling fixes
aniketmaurya · x · 2026-10-28
CI platform Celesto reports CI minutes grew 9x last week as AI-generated code floods pipelines, exposing bottlenecks hourly. What they shipped and are shipping:
- Compute: users run tests concurrently on tens of 8-16 vCPU machines; pool greatly expanded. Max concurrent jobs per org went from 20 to 50 in a week
- Queueing: Celery was hard to scale; affected tasks moved to NATS, cutting queue time by 20 seconds under 100+ concurrent load, with a full switch this week
- Security: applied all GitHub-recommended hardening protocols
- Context: a Blacksmith user complained about outages every other day while being billed per minute
A sign that CI/test infrastructure is becoming a real bottleneck in the AI coding era.
More from coding & agent
- Not Everything Needs an Agent: Fixed Steps Are Recipes, Save Agents for Judgment — adamse · 2026-10-28
- Dev suggests self-hosting via Tailscale only: no open ports, tailnet as sole access path — vincesitzmann · 2026-10-28
- Letta's FrontierMemory leaderboard: Letta Dream tops at 73.3%, beating OpenAI and Claude agents — afurgs · 2026-10-28
- jackson-databind hit with CVSS 9.8 flaw CVE-2026-110879, bypassing prior fix — Kyrannio · 2026-10-28
- Researcher says Claude writes 50K lines of code for him per month — ethanCaballero · 2026-10-28
- HaMo Combines Experience Cards with Model Adaptation for Long-Horizon Web Agents — Zhen Bi · 2026-10-28