After generating a basic program addressing the needs of the user, I examined the site, which was quite inhospitable at first, second, and even third glance.  Without any clear idea of what the building would look like, I began diagramming through 2d and 3d methods, starting from the program and using the topography and building site as reference points to push and pull planes and lines as a means to develop a form that was both referential to the site and functional for the user.

Because the site has a pronounced ravine running through the eastern half, I decided to employ a cantilever structure to float a portion of the structure above the ephemeral streambed.  Gradually a hierarchical system began to emerge, with a steel framed bent "ribbon" wrapping around the entire top and sides of the program, while a heavy stereotomic "counterweight" grafted the building to the site while also holding service areas of the program.

About halfway into schematic design, structural calculations informed me that I needed a significantly thicker cantilever root than I had predicted.  In order to accommodate the root while maintaining the ribbon's thin edge profile, I designed a system of tapered cantilever beams that start thick at a fixed point of the concrete counterweight, then thin out as they reach the edge of the ribbon.  A similar approach was taken with the cantilever roof structure, although the sloped area dies in where the glass envelope begins.

After an intense experience that eventually brought to fruition a complete  model made with real concrete and fabricated wide flange sections, I decided to flesh out the digital model I had used for the design process.  This allowed me to render "glamor shots" of the design, some of which you see presented here.