The poster has Claude act as an automation engineer, doing design and research work and finding related docs and indexes for their job.
The poster has Claude act as an automation engineer, doing design and research work and finding related docs and indexes for their job.
The poster had Claude write a Python script for a contracting job that extracts information from websites, because doing it manually was too time consuming. They use Claude as the decision-maker around that code: they tell it to start, it runs the workflow and returns the finished work, and the human mainly supplies the field expertise about where to look and submits the result.
The poster says it now covers what had been a 40-hour-a-week job and that their own involvement is down to about 10 minutes. They also say the LLM portion costs about 15 cents, but give no hard accuracy metrics or failure rates.
Their main caution is to avoid having Claude scrape directly; they trusted a script to do the extraction because it used almost no tokens and was faster.
Act as an automation engineer on my team. I need to automate [process, e.g. weekly report generation] using [tools you use]. Research the best approach, point me to relevant documentation, and produce a design with steps, inputs and outputs, and risks. Finish with a recommended plan I can review before building anything.
Score every demo request against your ICP, update HubSpot and ping the right rep in Slack within a minute of the form submit.
Every new Zendesk ticket gets a category, priority and sentiment tag plus a draft reply in an internal note before an agent opens it.
Copilot and Claude review each PR for bugs, missing tests and risky changes before a human looks, so reviewers can focus on design.