As far as I know auditd can't be configured to capture the file content. You probably wouldn't want it to anyway due to performance hit. So you need to create a baseline to which to compare the changed and original content.
If what you are auditing is a very small set of data you can simply use a frequent backup (or snapshotted filesystem on SAN or something) as your baseline version and do manual diffs. It seems, however, that you are trying to avoid manual solutions.
What it seems you are looking for is File Integrity Monitoring, just as @Michael mentions. These basically run a baseline scan and monitor for (a configurable set of) changes to a defined set of data. I say configurable, because you can include or exclude the file's content from the scan; it's all policy based. As the name implies, the purpose is to monitor the integrity of a system, not necessarily data content. So you will generally monitor application config files and compiled code, critical OS files, etc. to ensure that the system hasn't been compromised. These can also be used for network devices and databases and popular applications. Network devices are easy because the entire config is in a single file (or small footprint anyway).
We use Tripwire in my shop, where we have a great deal of compliance to deal with, but there are others: http://en.wikipedia.org/wiki/File_integrity_monitoring
FIM solutions will generally send alerts and reports to help automate the monitoring part. They will also have compatibility or partnerships with other solutions to integrate with SIEM solutions, even Splunk!. If you have a lot of monitoring to do then you'll spend all of your time reviewing logs. So, basically, "security is hard" ;-) These things are scalable and multiplatform and come in agent and agentless forms. You probably aren't the only one with such needs in your org. Any expense (and it might be less than you would think) can likely be shared to secure more than just your application.