I’ve been using ChatGPT alongside several other AI assistants for software engineering, and I wanted to share some positive feedback.

Over the past few weeks, I’ve noticed a significant improvement in how ChatGPT handles long, iterative projects. It no longer feels like it’s trying to answer each prompt in isolation. Instead, it follows the workflow I’m using, remembers the design decisions we’ve made within the conversation, and helps me build systems module by module.

For example, I’m currently building a production-quality document extraction pipeline. Rather than generating everything at once, we’ve been designing one module at a time, reviewing decisions, implementing code, summarising each module, and then using those summaries to continue in new chats. The experience has been remarkably smooth.

The biggest improvement is that it feels much better at understanding how I want to work, not just what I ask.

I’ve tried several other AI models during the same period, and ChatGPT has become my preferred assistant for large engineering projects because of this improvement.

Thanks to everyone involved in building it. Whatever changes were made here have had a noticeable impact on my day-to-day workflow.

Hey @CandyButcher, welcome to the community! Really appreciate you taking the time to share this.

It’s really nice to hear that ChatGPT has felt more useful for long, iterative software projects, not just one-off prompts. The example around building a document extraction pipeline module by module is especially helpful context.

This kind of real-world feedback is helpful, and I’m glad it’s been making a noticeable difference in your day-to-day work.

- Sunny