AI can help with software development in numerous ways, including operations, testing, and documentation. The list goes on.
At Pro-Sharp, we're deliberate about how we integrate AI into our daily work. We centrally control which tools our team can use. Every AI tool must undergo a security review before being implemented. No exceptions.
Recently, we've focused on agentic programming. The results? Mixed, but promising enough to keep pushing forward. This approach requires something different from our team - a new mindset and a different way of working. We gather experiences regularly, run workshops, and hold internal training sessions. The goal is simple: get the most out of the technology.
Here's the reality: agentic AI can deliver quick wins in the short term. But that's where the illusion ends. To ship production-ready code, you need senior developers, clear requirements, and solid quality assurance. That's where we're putting our focus right now.
Our experience with MVPs has been exciting. The results look impressive fast. But stable? Not quite yet. This is where experienced developers become critical - they know what it takes to build systems that actually hold up in production.
Some wins we've seen across coding and beyond:
AI has been particularly valuable in planning (turning business requests into structured specs), testing (drafting test cases and spotting coverage gaps), and documentation (summarizing complex modules and keeping internal knowledge accessible). On the coding side, it supports refactoring, code reviews, and feature implementation. These contributions make our team faster and more consistent, not just more productive.
What's your experience been in your field?