Skip to content

NASA · 2017–2025

ESSCOR

Catalog, search, and access software for Earth science holdings — metadata, discovery, and ordering in one workflow. There is no public demo.

Lead Software Engineer — catalog, search, and data-access software.

  • PHP
  • MySQL
  • Elasticsearch
On this page

Problem

  • Researchers need to discover data across holdings that did not share one metadata convention.
  • Discovery, access rules, and ordering have to work together, not as separate tools.

Decisions

  • Build a shared metadata model and catalog workflows.
  • Use MySQL for metadata storage and Elasticsearch for discovery.
  • Connect search results to granule-level access rules and automate dataset registration.

Evidence

NASA · 2017–2025

Outcome

  • A catalog that connects search, access, and ordering around one metadata model. There is no public demo.
  • An automated registration workflow so new holdings do not depend on a fully manual path into the catalog.
  • This page does not publish how much registration time changed. A percentage is not claimed here.

Team & contribution

Contribution
Technical development of a shared data catalog
Team & partners
Engineering, data-management, and research partners
What I improved
Manual dataset registration and disconnected discovery and access paths.
Key decision
Establish a shared metadata model before building search and access on top of it.

Catalog and search software for ESSCOR. The work brought Earth science metadata, discovery, access rules, and ordering into one catalog so a holding was not a separate search problem.

There is no public demo. What can be said is the system and the decisions, not a live artifact and not a time-saved figure.

Search on a shared model

MySQL stored the metadata. Elasticsearch provided discovery. A shared model made holdings easier to describe and query instead of giving each collection its own search path.

Granule-level access rules connected a search result to the access and ordering workflow that applied. Discovery without those rules would have stopped at a list.

Registration

A content-registry workflow automated the recurring steps that take a dataset from arrival to something the catalog can describe. The aim was less manual coordination between new data and a researcher being able to find it.

How much that reduced registration work is not published here.

← All work