The Central Role of the ML Model Owner for Enterprise Data Science
For AI projects looking to go live and deploy to an application/API, the focus is mainly on the AI product manager and technical lead. But someone needs to guide the model through the entire lifecycle
So you’re in an AI-naive organization. You’re trying to make your AI project go live and deploy in production, via an enterprise application or an API for consumption downstream. For small organizations, having the AI product manager in business partner with the technical lead of the data science stack should suffice to build complete AI/ML systems, end to end.
However, for large Enterprises, the large amount of employees, siloes, and bureaucracy can have many impediments/blockers to making your AI project go live, despite having an AI product manager in business and ML technical lead in IT. For large Enterprises, these two important roles that worked at smaller organizations, doesn’t work work so well when having more hoops/policies/gatekeepers to go through.
Sure, you have your use case, requirements, data, prototype, MLOps pipelines for deployment, and your frontend to consume it for your Enterprise users. You have the right use case, right data, right talent, and right technology. But somehow there are too many blockers where your AI project gets stuck in the various phases of the data science lifecycle, that it results in delays and the costs for your project increases with each week that passes with no progression. You can’t see the light at the end of the tunnel, and you are stuck in AI prototyping purgatory. In the meantime, other AI groups are trying to do your exact objective, which is to make their AI projects go live. And unfortunately, they run into the same problems, and they start duplicating the work already done in other areas, only to get the same result- their own AI project can’t go live.
The problem here is that while each part of the data science lifecycle is assigned to different areas, the vastness and bureaucracy of the large Enterprise makes the job of the AI product owner and technical lead not empowered enough to oversee and guide the ML model all the way through the DS lifecycle and navigate through all of the gates and siloes. I opine you need an ML Model Owner to join the leadership team led by the AI product owner, so that the ML Model owner can ensure the model flows through the entire lifecycle, and ensures the integrity of that lifecycle for an AI project to go live.
At large Enterprises, the AI project manager and technical lead are too high level to ensure the ML model at the project level is able to get through all the phases of the DS lifecycle. Someone needs to wear the hat of the ML Model Owner, and given enough authority at the Enterprise level to address the blockers at the gates between siloes to go end to end and make your AI project go live. Usually, this ML Model Owner will be the senior data scientist the is with the business line that funds and owns the AI project.
Stay tuned as I detail the specific duties of an ML Model Owner, and how they partner with the AI product manager and technical lead for making an AI project go live at your large Enterprise.
If you're looking for support from me, here are a few options:
Enterprise Data Science Consultancy: With my consult team comprised of a Senior Data Scientist, Senior ML Engineer, Senior Data Engineer, and Senior Cloud Engineer, we will help you architect and build your Enterprise Data Science platform, and transfer knowledge to your IT team to maintain and optimize it. We will also overlay an MLOps framework to manage the AI solutions you build on this platform. If you don’t have an MLOps team, we will help you build one. Please get in touch about this consultancy here
Coaching and Mentorship: I offer coaching and mentorship; book a coaching session here