This is my first bioinformatics job and my first return to an academic environment since finishing a MS in CS and doing the standard industry thing. This group is mostly a core services group rather than a pure research support group.
I came in to do traditional data warehouse work, loading data from separate databases into one spot to make reports across those different datasets easier. That's most of my days now - simple data integration and reporting. Oddly enough, we are an Oracle shop because our university has a site license. But I do a little Ruby/Rails/Sinatra for a project that is reporting + some app-like enhancements. Tons of SQL for manipulating data, but it is usually simple SQL, especially compared to stuff I wrote previously in a financial corp.
A significant amount of my day is ferreting requirements out of users. I wish I was better at this.
There is a much larger amount of grunt work happening in the data management space than I expected. We deal with a huge amount of protected health information (PHI), which can be frustrating since you're usually working around data policies that require somewhat careful interpretation. Obviously, the genomic datasets are huge, so dealing with storage and clusters for data processing is often a discussion point around here.
It is an academic/research place, so it is a little more relaxed than industry. Expectations (for software) are a little low - we're talking about users who have, for the most part, kept data in spreadsheets or little Access style databases for years. And they often haven't had the budget to hire dedicated software guys, so sometimes tools are suffering from bit-rot.
Good hackers seem to be appreciated. It is below your standard industry salary, but not terribly so. I try to avoid being sucked into low-value work since "the reward for work well-done is more work."
Just like grad school, if you want to work on a specific topic or research area, make sure you get a job in a group doing that work. Once you're established, it seems easy to get some space to do some exploratory hacking.