I work at a biorepository and we have around 700-800 extreme- and ultra-low temperature freezers (mostly -80C for mechanical and -185C for liquid nitrogen). They're monitored 24/7 by two separate servers running third-party software developed by GE/Kaye systems.

I created a solution that exported data from these two servers, imported it into an operations database in SQL Server and updated where necessary. It also serves as the data source for a Visio-based layout that we use as a map of our repository. The map displays the status of the units, their associated alarm points and will display further data about the unit or alarm tag by clicking on it. We also installed Visio Viewer on our workstations to display it in Internet Explorer so we saved on licensing costs.

Finally, it checks to be sure that units containing samples have their assocated alarm point turned on and emails a number of people if they aren't.

Prior to starting that project, I knew some SQL, but had no exposure to SQL Server. Needless to say I've learned a thing or three.

Right now I'm working on a solution to centralize production data that is currently kept in around 40 separate Excel spreadsheets (each staff person has one they fill out manually and the data is currently collected by using DDE in an Access application). Data consistency is, as you can imagine, a challenge. I'm creating a single application for data entry that will serve as the front-end for a SQLServer database. SSIS plays a role doing some additional data validation but eventually will include a component to transfer the data to a to-be-created data warehouse. That data warehouse is my personal Everest at the moment.

Our business unit collects a lot of data, but it isn't always the right data and it's not collected and stored in a structure that enables us to leverage it. And a fair amount of this data is mission-critical as in determining how much we bill our clients, labor costs, etc.