About Keith Binkly

I’m Keith Binkly. Over a decade as an analytics operator — Senior Analyst to Director at Green Dot, leading the team behind some of the largest Banking-as-a-Service programs in the U.S.

A decade fishing data the slow way

For most of the history of enterprise data, we’ve worked like 19th-century fishermen. Voyages planned over months, carried out over years. We ran the gear — trawling nets out, scoop up what we could, cast a line now and then to see what we’d catch. The tools helped, but we still pulled each fish off the line by hand and cleaned it with a trusted set of knives. More of the work went into finding, hauling, and cleaning the catch than into the part that mattered: the analysis, the patterns, the meaning.

There are whole oceans of data now. Most of it has gone the way the real ocean has — mostly unexplored, because our tools were slow and our reach was short.

That’s changing. What’s suddenly available is closer to a nuclear-powered submarine: it scans the abyssal plain, ingests the whole water column as it goes, processes it onboard, and throws a live readout onto the cockpit glass. The upgrade isn’t a bigger net — it’s that you can finally see where you are.

But the sub is still best piloted by someone who knows the point of the expedition. The years spent casting a single line at a time aren’t wasted; that’s how you learn the water. We’re just putting the fishing rods away now.

What I actually care about

What I keep coming back to is older than any of this. There’s a long arc of people finding better ways to understand the world, and the tools that widen it tend to arrive as step changes. The printing press didn’t improve access to knowledge — it broke it open. I think we’re in another one of those moments, and the part I can’t stop thinking about is visualization.

A good visualization is a portal. The way music carries across any language, the right view of a system lets someone grasp a phenomenon they’d never reach through a table or a paragraph. I’ve chased that moment my whole career — the instant a stakeholder stops squinting and actually sees what’s happening. But it’s bigger than work. The same thing happens whenever anyone uses these tools to understand something they care about, for no reason other than wanting to know.

So the question I keep returning to: now that AI lets us metabolize information in ways we couldn’t before, what new ways are there to see?

What’s here

This site is me working that out in public — a field library of what’s shaping my thinking, a running log of what it’s teaching me, and a lab where I test new ways to explore and visualize data. The log is written with an AI librarian that reads the library alongside me and synthesizes what’s emerging; the lab is where the agent-memory and context patterns behind my day work get pointed at something visual. Underneath all of it is the same job I do off the page — building agent systems that understand enterprise data in context, the kind of work that’s turned a pipeline I used to spend half a day on into under an hour. Some of it is enterprise analytics. Some of it is just curiosity. It updates as the work does.