I work as a developer contractor for Company A, where most of the developers and managers I know are not familiar with agentic development tools or what tools like Cursor and Claude can currently do. I also work part-time for Company B, where we use a highly automated workflow with specifications, reusable skills, tool integrations, cloud agents, and licensed AI development tools.
Before learning that workflow, my work for Company A usually took more than six hours a day. After applying the general ideas locally, I can now finish the same responsibilities in roughly 1.5 hours, including meetings, because I have automated much of my process.
I enjoy building these systems and would like to show Company A how guardrails, testing, code review, logging, and similar practices could help the whole team. However, I doubt sharing the productivity gains would lead to a better salary. I am also concerned that management might simply expect substantially more output and fill the time I save with additional work.
Would you introduce these practices openly, propose a formal pilot, or keep using them privately? Also, how would you handle any intellectual-property or confidentiality concerns when adapting techniques learned at another company?
3 Answers
The safest middle ground is to demonstrate one narrow workflow rather than unveiling your entire setup. Show how it improves quality or reduces cycle time, not just how quickly you personally finish tasks. Build the guardrails around ordinary human-written code too, since AI-generated code still needs review, tests, security checks, and clear accountability. Also separate general knowledge from anything belonging to your other employer.
I would not volunteer to create a large automation program for free. Introducing these tools could easily create more work for you: training people, handling failures, writing policies, and dealing with management raising everyone’s targets. Keep using your process to meet your obligations, but only expand it into a team-wide initiative if the company agrees on scope, ownership, resources, and a concrete benefit for you. Otherwise, the time savings can reasonably remain your efficiency gain.
I would be careful, but I would not hide it indefinitely. Do not frame it as “I can now do four times as much work,” because that may just become the new expectation. Instead, propose a small, measurable pilot focused on something like bug-fix throughput, review turnaround, or defect rates. Present it as a team improvement with code review, automated tests, logging, access controls, and spending limits. Before doing that, make sure you are not transferring any proprietary prompts, workflows, code, credentials, or licensed tools from the other company. If the pilot succeeds, use the results to negotiate compensation, a formal automation role, or protected time for maintaining the system.
Related Questions
Biggest Problem With Suno AI Audio
How to Build a Custom GPT Journalist That Posts Directly to WordPress