3 ms·
I make a lot of prototypes along with maintaining production systems. A lot of the projects have to do with ML/NLP, which basically means I have to chose python
by kubami 5y ago
I make a lot of prototypes along with maintaining production systems. A lot of the projects have to do with ML/NLP, which basically means I have to chose python as a core backend language. Which is great, I love python.
My core production system and every new prototype is:
- Django + PostgreSQL
- Celery + redis - for long-running tasks
- React
- GraphQL
- Terraform + Cloud-init to provision the infrastructure
All quite boring and stable, the way it should be :)
I have a longer writeup, along with my reasoning for the technologies I chose: https://kubami.com/articles/my-modern-saas-software-stack/ https://kubami.com/articles/my-modern-saas-software-stack/
- f6v 5y agoHow time flies... React, GraphQL and Terraform are considered boring and stable.
- kubami 5y agoFair enough for the GraphQL. However I'd say React is so widely use I'd say it is boring and stable (as the frontend tech goes) Terraform, well, what is the alternative? Clicking through web interfaces, or bash scripting the cloud-provider CLIs.
- axelroze 5y agoCould you share what python libraries you use in your projects for ML and NLP?
- kubami 5y agoSome of the libraries I use - nltk - preprocessing - spacy - preprocessing, entity recognition, part of speech recognition - gensim - Similarity and clustering - pytorch - ML model building