Engineering Blog
-
SAFE
How we stopped asking the model to write our UI
We let the model write our interfaces in a small invented language. The rules it needed were making the agent worse at everything else, so we took the language away and gave layout to code.
-
Future
The next wave of Enterprise AI will run on intelligence we control
Most enterprise work does not need the most powerful model in the world. Once good-enough intelligence runs in open weights, cost and control decide the architecture, and the hard part of agents is everything around the model.
-
People & Process
Introducing PodFlow, our AI-native R&D operating model
How Sidetrade R&D takes an idea all the way to something a customer can use: what PodFlow is, what was slowing us down before it, and where every human decision sits.
-
.png)
Platform
The agent is the easy part
First in a series on what enterprise collections agents need beyond the model: why an agent harness, not the model itself, is what makes an AI decision safe to run at scale.
-
People & Process
You are nothing without your team
Five days in Portugal, 120+ Sidetraders, 18 projects and AI in nearly every one of them. Why our annual hackathon is the most important week in the engineering calendar.
-
Platform
Why we run Sidetrade IQ on ourselves first
Sidetrade IQ is the internal deployment of the platform our customers get as Aimie IQ, on the same framework, the same models and the same controls. Why inside first is how the intelligence scales out.