The need for data science grey matter and building your talent pipeline in Enterprise settings
Enterprise Data Science, practiced at large organizations, requires grey matter. Without data science talent, you can’t shift from the legacy systems that weigh down innovation to AI enablement.
Many large enterprise organizations, especially government organizations, are weighed down by legacy systems. Some are decades behind AI-first companies, while others are so behind that they are forced to sunset systems that are no longer supported by the vendor.
Now imagine trying to start Enterprise Data Science capabilities in such organization. And I’m not talking about the AI prototyping and local AI application ‘deployments’ on local environments- I’m talking about building an Enterprise Data Science Platform. These are complete ML systems at Enterprise, where you deploy AI/ML models in production, at scale. In other words, these are Enterprise end to end AI applications, where there is collaboration between business, the data science organization, and IT to build these Enterprise AI applications within an MLOps framework.
We are talking about Enterprise Data Science, and to start the shift towards this target state, you need data science grey matter. And you need numbers. Let’s face it- as one of the few data scientists in such an organization, you are outnumbered by all the data analysts, statisticians, database administrators/developers, ETL developers, data modellers, BI specialists, and software developers. While they may be good with DevOps and software development and BI dashboards, they don’t have a clue what MLOps means, and how to maintain the integrity of the ML development life cycle throughout the process of building end to end AI solutions.
As a trailblazing data scientist at your enterprise organization, the key to building data science capacity is hire more data scientists. And you need several of them to start, as you need to help define AI governance, AI development guidelines, AI standards of practice, AI project intake, ML reference architecture, and an MLOps vision. More importantly, you want to define how data scientists work in your organization, and define a career path for them.
Once you get these foundational documents/blueprints/frameworks in place, then you can start building a minimal viable platform, a Strawman, where you can iterate on the best architecture to accommodate your AI use cases that pass through it. And if this is cloud, you can quickly tear that Strawman down and quickly rebuild according to what you need with your use cases that passes through. Soon your Strawman becomes your Enterprise Data Science Platform.
But you need data science talent to get this massive, multi-year foundational work done. One way to get talent is to build your data science pipeline via data science graduate programs. I managed to get half a dozen data science professionals hired at my organization from my connections with my alma mater, from the university where I completed my Masters degree in Data Science. When I graduated, I kept close ties with my school as an alumnus, and developed networks with the professors who taught me in grad school. Through these graduate school networks, I managed to meet many of the students and graduates of my university. We initially started by hiring their data science students as interns, and then started to hire these interns as permanent employees.
Thus, this was the start of a data science talent pipeline, which continues and provides my organization with much needed data science talent. Because without data science grey matter, all you have is talk. At the end of the day, you need talent to code and build your complete ML systems at 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/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. If you don’t have an MLOps team, I 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