I will take your models to the next level making them production ready with mlops


About this gig
Have you or your data scientists developed a wonderful ML model you want to show to the world? show to your clients? let them use it?
Or have you even developed some model, and then worked on a new version of it, only for your client to tell you "show me again the version we saw on Saturday" and you are lost on what version since you have moved on?
Or do you excel at working your ideas on a jupyter notebook with nice interactivity but find it bothersome to write so much "filler in" code?
Wouldn't you like to have your model transformed into a pipeline, and perhaps run experiments with it with different parameters and show them?
This is where I come in!
I am a professional ML engineer with years of experience and passion for this term in every person's mouth: MLOps. My background is a Software Developer so I can contribute to your success and your data scientists can concentrate on what they do best: Train models while I build the pipeline and add functionality to it to be in production.
The Basic, Standard and Premium packages offer different stages, contact me for clarification
Tools: Python, Poetry, Tox, Plotly, Pydantic, MLFlow, DVC, Hydra, scikitlearn,etc
Get to know David A
AI Software and MLOps Engineer
- FromJapan
- Member sinceAug 2024
Languages
English, Spanish, Japanese
FAQ
What does the Basic Package offer?
The Basic package transforms your notebook into a python package with configuration files and input validation. (Tests can be added for a additional fee). I usually use Poetry but other ways can be arranged
What does the Standard Package offer?
It offers to add either experiment tracking (with MLFlow) for your model pipeline, or Data Versioning in case your data changes frequently. The final specs can be arranged after consultation.
What does the Premium Package offer?
Here, I can provide an API with FastAPI so that your users can actually try your model and see the results. Authentication can be added on request after consultation
