Virtual LinkML community workshop — gauging interest

[matentzn]

Floating an idea: a virtual workshop that brings the whole LinkML community together to show what we're each doing — schemas, tools, development workflows, anything you've built or are building with LinkML.

The point is simple: learn how others go about things. There's a lot of LinkML work happening in parallel across very different domains (biomedical, public sector, geo, environmental, library/archive, …) and most of us only ever see a thin slice of it. A workshop is a low-cost way to make the rest visible. And it could just be fun.

Some ideas how to structure this:

Format

• Fully virtual
• One day, possibly two if there's enough interest
• Talks based on 1-page abstracts
• Light structure: short presentations + Q&A, with thematic sessions if the abstracts cluster
• Recordings posted afterwards (with presenter consent)

Who can present?

Everyone. Schema authors, tool builders, downstream consumers, occasional users — academia, public sector, industry, hobbyists. No filter.

One prerequisite: whatever you present needs to be openly available — a public schema, a public repo, a public tool. It doesn't have to be polished, production-ready, or actively maintained; people in the audience just need to be able to look at it afterwards.

What makes a good talk?

Anything you find interesting. As starting points:

• A tour of a schema you've built — what's in it, what was hard
• A development workflow you've found useful (CI, testing, doc generation, releases)
• A tool you've written on top of LinkML
• A migration story (e.g. moving from another modelling language to LinkML)
• Lessons learned, dead ends, "things I wish I'd known earlier"

Live demos welcome and encouraged. No need for slick slides.

Interested?

Just add some emoji of choice or a comment with ideas.. We will take it from there.

[caufieldjh]

I'm curious if the scope could include "I want to do this in LinkML but don't know where to start" or even "Is this a use case LinkML has tooling for already"

[matentzn]

Great Q! I don't know yet - up to us I guess!

[StroemPhi]

I'd love to present on our DCAT-AP+ and ChemDCAT-AP work to get more feedback from the LinkML community.

[matentzn]

Super cool, I agree!

[gouttegd]

Happy to present how we plan to use LinkML for imaging metadata.

[clevinson]

Seems like a great venue to share some of the LinkML powered agentic data extraction (+ data auditing app for verification of LLM output) that I've been developing. Consider me interested!

[ddooley]

I could present DataHarmonizer 2.0 beta as well as a related command-line app currently called Menu Manager for managing enumerations whose source exist in other agency products - with modules parsing html pages, pdf, owl, skos, json and API sources for AGROVOC, STATSCAN, USDA & Canadian soil database documentation, SNOMED, and others.

[diatomsRcool]

I would present on microschema! Or maybe the exposomics model

[dhruvrajan]

I'd love to attend, and could aim for a presentation on ontology-powered data analysis.

[matentzn]

Maybe invite these folks on schema editing interfaces: https://github.com/orgs/linkml/discussions/3478