Managers focused on day to day data center operations in bigger organizations tend to develop an ambivelent attitude to open source - it's important, but irrelevant.
One way to bridge that divide is to think about using open source as a recruiting tool. Look carefully at your development plans for the next year or so and ask what could be be generalized to the point where it would gain value for others, without losing its value for you. Then ask your staff the same question -without restricting them to the approved list.
Notice that developing something as an open source project has vastly different implications than the 80s idea of monetizing your intellectual property assets by selling applications - that generally didn't work out well initially because most of the products offered were of poor quality and in the long run because even the successes turned out to be unmaintainable within the internal systems organization -leading to a few successful spin-offs and a lot of recriminations.
Do open source right and none of that should happen. First, you spin off liabilities before you even begin: support your staff in creating these as personal projects to which you donate corporate resources -including test users.
Get solidly behind the people involved and you'll soon have several projects going - and once that happens you can start to use their presence in your group as a recruitment tool,: simultaneously boosting the average quality of the people you recruit, job satisfaction in your organization, and whatever business purpose the project, or projects, were supposed to help along in the first place.
It's true that those who succeed will eventually leave to pursue wider commercialization or other interests - but you'll have a successful project, the people involved will typically stay with you longer than they otherwise would, and both you and your organization will likely have earned their loyalty for life. It's the ultimate win-win.