Last week, I shared some lessons learned from a Domino Data Science Pop-up that I attended a…
Data / AI / ML
A few weeks ago, I attended Rev, a summit for data science leaders and practitioners organized by Domino. I have previously shared blog posts on the key takeaways from last year’s event (which was one of the best data science conferences I’d been to), including how to define a data scientist and the differences between data science and data […]
Last week, I shared some lessons learned from a Domino Data Science Pop-up that I attended a…
A few weeks ago, we hosted our second marketplace meetup in San Francisco. The night…
A few weeks ago, I attended Rev, a summit for data science leaders and practitioners organized by Domino. I have previously shared blog posts on the key takeaways from last year’s event (which was one of the best data science conferences I’d been to), including how to define a data scientist and the differences between data science and data engineering.
This year, one of the core themes was the management of data science projects. I had never thought about it this way before, but in fact, models are the output of a data scientist’s work.
My biggest takeaway was the realization that when it comes to building products and businesses, it’s important not only to be data-driven, but also to be model-driven.
What does it mean to be model-driven?
If data is oil, then the model is the engine. Data, and code for that matter, can be unitized, whereas models are more complex. Models are effectively created with data and code as building blocks. As a result, models can’t be treated in the same way as data or code, and there are definitely best practices for managing the data science lifecycle.
Generally speaking, the overall data science lifecycle can be viewed in an aggregate of stages:
Most of us likely have intentionally plotted just a few of these stages. While it might take a lot of work to put a comprehensive workflow like this in place, the biggest benefit is that we can truly measure everything. This hopefully translates to greater efficiency and more effective collaboration between all stakeholders, with more auditability and reproducibility.
If you’re interested in more details on building a model-driven business, I recommend checking out Domino’s whitepaper on the topic.
Data / AI / ML
This November marks three years since the launch of ChatGPT. That moment brought AI into the mainstream, with large language models (LLMs) seen as the breakthrough technology powering it. Since then, innovation in AI has been relentless — perhaps one of the fastest cycles we’ve ever witnessed in tech. It’s worth pausing to reflect on […]
It’s hard to believe that it has been three years since my first day at…
“It takes 10 years and $30m to become a great investor.” This quote has stuck…