Many places if you want to write software that is close to the action , performing the computations and correlations that discover new things...then your best bet is some sort of technical credentials, some sort of field credential (get a degree in biology or find an internship/job somewhere) and then find a job in a university or industry lab which is working on something cool.
Then there are the citizen science type projects which take data from NASA and various other places and present it in a more useful manner so new things can be inferred - like Galazy Zoo for example.
Another option are likely to be small independent organizations that are working on smaller research schemes and need help in basic day to day stuff e.g. tools for automatically pulling citations out of pdf files and linking them together for paper discovery or simply just cataloguing and presenting know papers in an easy to digest format for public education. These may seem relatively small but every little bit helps push the boundaries, especially in some underfunded areas.