Over the past decade, real-time control design has rapidly moved from the age-old process of developing code based firmware in embedded C, to platforms that auto generate the embedded C program from a higher level drag and drop, model based design. This has various advantages such as device independent system design, reduced possibility of user based code translation errors, limited effort needed in terms of code maintenance and most important of all, ability to design more complex systems without worrying about lower level implementation. Various softwares are commercially available such as Matlab, Plexim, PSIM, VisSim etc., for this purpose but they all have the same shortcoming, i.e., extremely unaffordable to individual users and small companies. In this thesis, a new numerical simulation platform has been developed from scratch. To this end, a new programming language with inbuilt matrix support was created. The implementation details of the various compiler components, such as lexer, parser, sematic analyzer etc., has been presented. In addition to this code based platform, a model based platform was developed, to enable drag and drop based system development and, the design details including solver development, tools supported, real-time controller peripheral support etc., are presented in detail. A common platform independent intermediate language was created, to which all model and code developed in this platform gets translated into. Multiple up compilers are made available with the platform to enable translation to other languages. One such supported language is C89 which is the most commonly supported language by real-time controller manufacture’s standalone compilers. The auto generated code adds all necessary peripheral and system configuration code and control code, compiles it to device specific hex code and programs it directly into the controller. This allows users to have no knowledge of the underlying device or any programming and makes the entire system more readable and easy to maintain. For optimized running of the generated code, a high-performance math library was created using minimax polynomials and implemented directly in supported device specific assembly. In addition to this method, other commonly used methods are analyzed in depth. In addition to this, various other forms of optimizations were implemented that are not available on other platforms. In the current version, only TI’s TMS320F28335 is supported with plans to extend it to more controllers. Two low cost rapid prototyping platforms were developed to enable real-time control using model developed in simulation. The solution was tested on multiple motor control and power electronics applications, the results of which are presented. This whole solution cuts down the cost of numerical simulation and rapid real-time prototyping to few thousand dollars which would have previously costed upwards of ten to hundred thousand dollar on a platform of similar capability.