OneForma's 6 Project Categories Explained

OneForma's own jobs-page filter has exactly six official categories. Here's what we could verify about each, based on real listings we found and opened.

1. Annotation

The largest category by real examples we found — includes AI response/rubric (a scoring guide that defines what counts as a correct answer) evaluation (Project Meteor, covered in our [dedicated guide](/articles/oneforma-project-meteor-guide)), independent QA review (Lighthouse QA), and medical annotation (labeling data — photos, audio, or text — so AI can learn from it)/report generation (Project Slice, US-only, aimed at radiologists and medical students/recent grads).

2. Transcription

Long-form audio transcription and speech annotation — the Freya Project, which we cover in full [here](/articles/oneforma-freya-transcription-guide), is the real example we verified.

3. Judging

We observed three named projects in this category on OneForma's jobs board — AI language specialist work (Project Hermes), speech-translation data quality evaluation (Project Zeno), and AI safety/red-teaming evaluation (Project Harbor Red Teaming – Juno) — but did not open their individual full requirements in our research pass. Titles are confirmed real; details are not yet verified.

4. Data Collection

We observed one named project — ego-centric (first-person) video data collection (Project Triangle). Title confirmed real; full requirements not opened.

5. Translation

Listed as an official category in OneForma's own job filter, but no live listing was open in this category at the time of our research.

6. LLM Prompt Authoring

Also listed as an official category with no open listing to verify against at the time of our research.

Final Thoughts

OneForma's categories are broader than the three projects most guides (including most of ours) go deep on — worth checking the live jobs board directly for Judging, Data Collection, Translation, and LLM Prompt Authoring listings, since availability rotates.