First Contribution by Meggan Nyakwaka

[MegNyakwaka]

Name : Meggan Nyakwaka Project: BentoML Governance Document: BentoML Governance Model

It is important to understand first what the organization does in order to clearly decipher whether the governance model implemented actually serves every party involved. For a beginner, to both open source and Machine Learning which is what BentoML works around, I had to take a few moment to get the general understanding of the concepts behind this organization. This was from general notes from Geeks for Geeks and resources from the opensource website.

Let us begin with some small background about BentoML. It is an open-source platform that provides a framework for serving, managing, and deploying machine learning models. It is more especially useful to Data Scientists and Machine Learning Engineers. For context, this tool helps them where they intersect which we call model serving. Data scientists have the expertise to building the model and training it. The ML engineers on the other hand have the skills to deploy the models but not the ones that require continuous improvement. So BentoML acts as the go between to ensure seamless working of data scientists and the engineers in the entire process and ensure that whatever improvements made to the models by the scientists does not affect the stability of the end product on the engineers side.

Governance Model: As it is an open and community driven project, anyone with interest can join, contribute to the project and even participate in the decision making process thus it takes a consensus based structure. It is however also meritocratic whereby the participants gain influence over a project through the recognition of their contributions.

Project Roles: Users Contributors Core Team Project Lead or the Benevolent Dictator

The Decision Making Process BentoML uses a Lazy Consensus decision making process which has the following steps: A Proposal is put forward followed by a discussion which they typically give an allowance of 72 hours for all the community members to contribute to the decision. At this point if nobody opposes then it is assumed that everyone has agreed to the implementation of the proposal. If a consensus is not reached in the discussion, then the proposal is moved to voting where the community is encouraged to state their opinions in the discussions and votes. In the end, the core team will have the binding votes at which point the decision will be made. Their mission and vision revolves around simplifying the deployment of machine learning models and make it accessible to everyone. They aim to achieve this by providing an open-source platform that can be used by developers and data scientists to package their machine learning models into a production-ready format, which can be deployed to any cloud platform or on-premises infrastructure. With this is mind, I would say that the governance model especially the decision making process which is what strongly steers the direction of projects works as every community member is involved.

As a beginner, it was quite easy to locate the documentation as it was directly in their GitHub repository. It was also easy to comprehend as it uses simple English and has several relevant links for more information especially on guidelines for the public which makes everything very elaborate.

[MegNyakwaka]

@arliss-NF Could you take a look at my contribution? Looking forward to your comments. Thanks!

[MegNyakwaka]

Update: I had a few thoughts as I was working on my second contribution and made a few edits here, I hope that is still okay @arliss-NF

[Chimnwendum]

Hello Meggan, it's great to have you here.

In the first contribution under the roles section, we were asked to just list the roles.

Also, there is nothing wrong with making corrections to existing contributions. It shows you are learning.

All the best.