If you're looking for my resume, click here: pdf | html

To see my current side project [cuomo], click here.

To see my undergraduate research, click here.

To see my undergraduate AI project, click here.

University of South Florida - Academic Computing - December 2000 to August 2003

FACT Reference Material Database
Database of all the products that we support in the FACT Lab. It includes descriptions, screenshots, sample files, related-links, as well as a mini-message board that allows faculty members to post their thoughts and opions of the various products.

Training Workshop Database
A package of various modules that allow the trainers to modify the training workshop schedule without actually having to edit any html. There are other modules that allow the addition of classes, categories of classes, and trainers. The current list of classes and trainers is used to drive the schedule creation module. There is also a statistical module that pulls various stats from the database such as the least/most popular classes overall, most popular classes between certain times, and there will be more stats to come. Information from this database is also used to dynamically create the Workshop Request Form -- so any additions, deletions, or modications to the classes (or categories of classes) will be automatically reflected on the request page.

Online Trainer Evaluations / Online Guest Lecture Evaluations
All evaluations are stored in a database so that various stats can be pulled. Having a database of constant feedback will allow us to continually improve our training workshops.

Workshop Request Database
All requests for workshops are now logged in a local database. We had previously sent all requests to the RightNow Web Q/A system as a means of logging all requests. However, by storing all requests in a local database, we can pull statistics that can be used to automate the workshop scheduling process.

If you have any desire to see one of the things that I've worked on at Academic Computing at the University of South Florida, click here. I am responsible for all of the links on the right side of the page. All of the links (except the first two) take you to pages that are created dynamically by PHP. Virtually all of the information that you see is pulled from a MySQL database on the fly. There are several administration pages that allow the content to be modified via a web front-end rather than a manual manipulation of the html itself.

typedef.org

what's been done:
- session variables for user_name, user_id, and class
- admins can now provide links in their posts
- add "Title" input box to post form
        - test for title; display "[No Title]" if no title (provides reverse compat. w/ old posts)
- enable delete links
- log invalid password attemps; log user applications
- admin public / private announcements (message bases 1 and -1 respectively)
- create dynamic links by pulling from several places (eg. slashdot, the onion, ?) w/ a cron job
- add web interface to add message bases
- slashdot articles are now javascript pop ups
- decrement post count on user stats when posts are deleted

to do:
- continue to go through code, refining functions (eg. is_admin can be phased out now w/ $HTTP_SESSION_VAR['class'])
- add interface to change login message
- message to next user
- more dynamic everything (random quotes, links, ???)
- make things more configurable
	- colors
	- real name posts vs username posts
	- min class required to see a particular message base
- refine message base admin page
	- change all the message_base_ids of the messages in a deleted base to the id that represents our "junk" base
- refine delete links for messages so that when individual messages are deleted, they go to the "junk" base
- refine delete links for users so that "deleted" so that they actually just disable their account
	- an admin should be able to then delete a user permanently from the disabled list of users menu

NEW to do's:
- allow users to edit their posts

and yet there is always more...

write a game that learns
- determine the appropriate combination of learning techniques
- possibly start with a statistical learning model