Research
Some technical choices are difficult because the technology is new. Others are difficult because the market is repeating a simple story that does not hold up under closer inspection. We research the mechanism, the economics, and the architecture underneath it.
Current research
Bolt-On AI is often treated as the weaker alternative to AI-native software. That framing mixes architecture, delivery, pricing, and marketing terminology. This research looks at when extending an existing platform is strategically preferable, what economic value it can capture, and how deeply AI has to interact with the system to create something new.
How data, permissions, tools, evaluation, workflow orchestration, and adaptive interfaces change the architecture of existing products.
Why faster code production does not automatically create faster product delivery, and what has to change around it.
How we research
We use primary sources, product documentation, market evidence, technical analysis, and direct operating experience. The goal is not to predict everything. It is to make a consequential decision with fewer untested assumptions.
Many debates stay confused because architecture, business model, delivery method, and marketing language are treated as the same thing.
A feature description is not an architecture. We look at the data, actions, controls, workflows, and operating model underneath it.
The useful question is not whether a system uses more AI. It is what the architecture enables the product and business to do.
Start a conversation
We can run focused research, compare technologies or vendors, and turn the findings into a decision the team can execute.
Talk with NewGeometry ↗