We built Engine with the belief that a lot of backlog type software engineering tasks could be handed off to a remote AI agent and completed async with little-to-no human involvement.

Recently, a wave of new models and products has attracted a lot of attention to async coding agents so I wanted to share our approach!

- To solve real engineering tasks for folks working on 'serious' projects we integrate tightly with Linear, Jira, Trello, and ClickUp and also both GitHub and GitLab.

- We quickly realised that configurability and workflow was important so there are lots of ways to set up and interact with Engine.

- Environment set-up is a pain at the best of times so we built an agent that helps out with this.

- Engine runs on all frontier LLMs, including Claude 4, which seems increasingly important with the pace of progress and competition amongst the big labs!

The free tier includes Claude 4 sonnet so would love for you to take Engine for a spin and let me know how it goes!