Cleo
Find and screen specialists for a defined program.
Open Cleo
AuraOne / Human Data
Qualified people create, review, and improve the data AI systems use.
Qualified people create, review, and improve the data AI systems use, with training, rights, quality, and delivery evidence attached.
AI teams that need specialist-created labels, rankings, evaluations, speech, demonstrations, or reviewed examples.
- Input
- A defined task, required expertise, data rights, and quality rules.
- Output
- A scoped dataset or evaluation record with quality and delivery evidence.
Capabilities
Enter through a concrete job and decision.
Workforce
Build and operate a qualified specialist bench.
Open WorkforceAnnotation
Create, rank, correct, and review training examples.
Open AnnotationVoice
Collect consented, license-scoped speech and voice data.
Open VoiceRobotics
Managed robotics programs collect and review human demonstrations, then coordinate a monitored handoff; queued transcode, copy, or webhook work is not represented as delivered.
Open RoboticsAura Capture
Join smartphone and wearable human-demonstration capture programs.
Open Aura CaptureFrom input to finished Human Data output.
Input
Bring the work and requirements.
- A defined task, required expertise, data rights, and quality rules.
Work
AuraOne runs the workflow with clear ownership.
- Recruit, qualify, train, assign, review, adjudicate, and document.
Output
Receive a usable result and complete handoff.
- A scoped dataset or evaluation record with quality and delivery evidence.
Operating flow
Each checkpoint moves the work forward with a clear owner, review, and next step.
01
Define the work
State the task, eligible expertise, source material, intended use, rights, and acceptance criteria.
Program brief and rights scope
02
Qualify people
Match specialists to the work and retain qualification, identity, consent, and training records.
Qualification and consent record
03
Create or review
Assign work with visible instructions, rubric, compensation terms, and escalation paths.
Task, rubric, work product, and timing
04
Check quality
Review examples, adjudicate disagreement, record feedback, and route rework without hiding failures.
Quality review and adjudication
05
Deliver
Package accepted work with source, rights, quality, version, and handoff records.
Delivery manifest and acceptance
What you receive
Human Data program deliverables
Your program keeps source material, criteria, decisions, approvals, and final outputs together.
- Program
- Purpose, task, owner, expertise, intended use, and scopeApproved program brief
- People
- Qualification, training, identity, consent, and roleContributor and reviewer records
- Work
- Source, instructions, rubric, output, timing, and escalationTask and work-item history
- Quality
- Checks, disagreement, adjudication, feedback, and reworkVersioned review record
- Delivery
- Accepted files, rights scope, version, checksums, and ownerDelivery manifest
What your team can do next
Connect the completed work to the decision, handoff, or release it supports.
| Outcome | Work | What you receive | Program fit |
|---|---|---|---|
| Specialist work is fit for review | Qualified people complete a defined task under visible instructions and terms. | Qualification, task assignment, rubric version, work product, and escalation history. | Best fit: programs with explicit expertise, task, and review requirements. |
| Data is accepted for the agreed use | Reviewers apply quality criteria, adjudicate disagreement, and record rework. | Review results, adjudication, rights scope, acceptance, and delivery manifest. | Delivery includes the agreed dataset, use, quality criteria, and acceptance details. |
| Program history is retained | People, work, rights, quality, and delivery remain separate but linked records. | Immutable events, versions, source references, owners, and final handoff. | Retention is configured around the customer's data policy and program needs. |