What we did
- With our skeleton bot set, we were so inquisitive to experiment with it. We tried various things with it and made it respond to most events - from a comment on an issue to the creation of a PR. We analysed all the payloads that the bot responded with to have a clearer picture.
- We explored deep into nlp-toolkit npm and had started to make our data set
- We later found out that npm natural would be a better fit for our use case and began exploring that.
- We had a full-fledged team call with all 10 members (including ourselves, all mentors, all coaches and the supervisor)
- In the meet we discussed about our progress and were advised to not start with NLP but build a simple bot with string comparisons to start with. NLP could be an optimisation we can work on if time permits.
- And yes, we sent out our blog post PR. Yay! :D
What we plan to do
- With NLP being postponed for later, we hope to start quickly with the implementation of the bot.
- We've chalked out incremental tasks to code the bot up and will get on it from tomorrow
- With little knowledge about NLP, we were having a hard time understanding the flow of the NPMs inside out, but now we'll get straight to the implementation of the bot ! :D
You must be logged in to add a comment.