There are lots of resources available, but I think following a design process might really help you most at this point.[1] You're going to need to conform the the style and practices of of the larger piece to which your part will fit in. Good design has to be consistent so that the user can have a continuous understanding. From this perspective, might it be possible to generate "working art" and approach the final visual treatment later? Also, for "beautiful" design to shine, usually this involves the removal of all extraneous elements, since good design is really about distillation.

Rough Design Process [2]

1. Discover

- Have you identified your users and the need that you are hoping to solve with this "part/feature"?

- Usually users are broken into different segments, a persona is just like a rough representative sample of the different segments.

- It sounds like you work for a for-profit organization, therefore you must also understand who your stake holders are and what they think the business needs are (they can be wrong and you might have to help correct this).

- Are there already people doing something like what you're trying to do? Not only should you benchmark competitors, you can look for analogous situations in totally different domains and still use this to help inform you.

- Redefine the problem your own way.

- review research claims, scientific literature, external info

2. Ideate - explore insights

- design objectives

- develop user POV and need statement

3. Create

- low fidelity prototypes based on the perspective of your user needs and POV from (2), explore solutions to the problems

- you should be going for volume at this stage because this is a discovery oriented process for finding unknown unknowns

4. Test

- escalate the fidelity of prototypes enough that you can gain feedback from potential or actual target users

- end points for higher resolution prototypes should definite measurable analytics and success metrics so some prototypes can be selected for further refinement, ideation and testing

- you can conduct formal usability testing, but the point of a prototype is to create an object to fuel conversation and discovery

[1] "1 hour of research saves 10 hours of development time" http://bokardo.com/archives/1-hour-of-research-saves-10-hour...

[2]This is my derivative of Stanford's Design Thinking process taught at dschool.stanford.edu, for a worksheet example see: https://static1.squarespace.com/static/57c6b79629687fde090a0... on dschool.stanford.edu)