Model in the browser and then send the model to the cloud for the computation, wait for some time, while our ever-growing back-end infrastructure brings you the results.

• GUI for modelling.

  • WebGl options:
    • Three.js
    • Explore the X-toolkit https://github.com/xtk/X

• Allow piping with python code and STAAD Pro

  • For python, following can be checked out:

    • ZeroMQ - maybe complicated, but very generic,and also ZeroRPC, and some info here (http://ianhinsdale.com/code/2013/12/08/communicating-between-nodejs-and-python/)
    • edge.js - Allows running python, .NET code from node.js
    • node-python

  • Verify results with STAAD-Pro, using .NET -> may be more robust than VBA

  • Include libraries like annyang for speech recognition

• Explore options for speeding up the server,

  • Try deploying parts to multiple services
  • Try map-reduce etc.
  • make the code non-blocking