To help developers reduce the time they spend in searching for syntax and code help from the internet and to save time spent on typing long and repetitive code multiple times. An application which directly returns the corresponding correct code based on the user's text prompt.
What it does
Takes text input from the developer/end-user which holds the code(syntax help) they are intending to receive.
How I built it
Text from the user is sent to Wit.ai. Here, the intents and entities are extracted and sent back to the web platform. Here, based on the intent and entity returned, we form and return the correct code statements to the user.
Challenges I ran into
Interactive Code Editor interface on a web platform. Overcame this by using a new Python web framework called streamlit.
Accomplishments that I'm proud of
Teaching many people how to use a fascinating platform like Wit.ai to build state of the art chatbots.
What I learned
Learnt a lot about NLP in general. Specifically more about intent recognition and entity extraction.
What's next for DevCoder
Adding support for many more programming languages and writing this tutorial in other languages.