Software Engineering

Multi-platform

  • Can generate executable binary files for different platforms and operating systems
  • Compiles to WebAssembly (WASM)
  • Transpiles to JavaScript (GopherJS)

Compatibility with different platforms (unofficially).











Common use cases

  • Command Line Interface (CLI) tools
  • Microservices
  • Web servers
  • Backend code
  • DevOps/MLOps
  • Glue code










Can Go be used for Machine Learning?

Yes, but it depends for what. Python is the most popular language for Machine Learning.

Several reasons for this, but the main one is the number of libraries and frameworks available for Python and lack of basic support in Go for e.g. CUDA and math/statistics libraries.











Go is powerful when it comes to preprocessing, data cleaning, and manipulating large amounts of data.

It is in large due to its speed and concurrency model:

  • High concurrency
  • High performance
  • Suitable for real-time AI applications
  • Suitable for data processing, big data, and data streaming
  • Great cloud support through libraries
  • Efficient use of resources, fast cold start










ML Go specific libraries

  • Natural Language Processing (NLP) go-nlp
  • Computer Vision gocv, with CUDA support

gocv











Libraries and frameworks

Some inspiration.