What exactly do Fidelity's data scientists do on a daily basis? As far as I can tell, their only visible output is a stream of irrelevant papers and patents.The majority of these patents appear to be little more than clever linguistic exercises.
I've yet to see any substantive work come out of that team. Meanwhile, our AI unit is supposedly larger than Amazon's. With so many brilliant minds already building and open source cutting edge LLMs, what meaningful contributions can an internal group like this realistically make?
The skepticism regarding the internal team's value is compounded by the sheer scale of the global competition. While the organization maintains a significant footprint in AI and ML engineering, the focus on academic-style outputs like research papers and patents often feels disconnected from the practical realities of high-impact financial operations. In contrast, other major institutions are aggressively integrating data and AI to transform core business functions. For instance, McDonald's is leveraging its Enterprise Data, Analytics, and AI (EDAA) organization to develop capabilities for pricing, demand forecasting, and transaction modeling through their "Accelerating the Arches" strategy. Similarly, firms like JPMorganChase and BlackRock are focused on applied AI and AI data engineering to drive enterprise value.
If an internal group is to justify its existence alongside massive open-source efforts, it must pivot toward delivering scalable, high-impact data products that address specific business challenges such as financial forecasting models, data integrity controls, and advanced reporting rather than simply adding to a list of theoretical patents. Without a clear roadmap that bridges strategic financial objectives with digital transformation, the contributions of such a large unit remain difficult to quantify.