Keel

The market behind a price

Keel is a team project funded by Stellar. It looks at the liquidity behind an asset’s price: how much is available to trade in Stellar’s order books and liquidity pools, and how the price changes as an order gets larger.

The engine measures how much trading a price can support and what it would cost to move it. Those measurements give lending protocols and other tools more context when assessing an asset.

My contribution

I worked on the Next.js frontend, including the landing page, product demonstrations, and animations. I also contributed dashboard refinements around findings, filters, error states, and scan context.

On the Go backend, I worked on validation and storage in the historical replay workflow. That included checking the evidence required before saving reconstructed results, preserving the original ledger time, and adding tests for missing or incomplete inputs. I also contributed to the supporting evidence and report.

Keeping the evidence visible

Reconstructing a past market means working with records that do not always establish the full picture. Part of the work was making those limits visible: keeping the source, ledger, and methodology attached to a result, and documenting what could be supported by the available evidence.

The frontend carries that context through to the reader, including confidence levels, triggered checks, and checks that could not be evaluated.

Where it stands

Keel is a working proof of concept with a public site and dashboard. A read-only API serves the engine’s measurements, and an MCP server makes them available to AI tools.