Limited published evidence · Archaeological research and rock-art documentation
RA3I’s described photograph-to-tracing documentation workflow
An affiliated author’s account and RA3I’s own documentation describe the workflow. They do not establish an inspected project, benchmark, privacy audit or customer deployment. The separate proposed rehearsal is designed and unexecuted.
RA3I’s published rock-art documentation workflow
RA3I platform described by Seth Chagi
Research/platform-described workflow published 13 November 2026; primary platform documentation corroborates review and export descriptions, not an observed project run.
A published third-party example, not a Nuvepro customer case or evidence of a partnership.
The workflow at a glance
RA3I describes a photograph-to-reviewed-record workflow
Illustrative summary of the published account
- RA3I image processing
Profile a field photograph
The published method describes examining lighting, surface and pigment to guide processing of the supplied photograph.
- NextRA3I processing workspace
Produce a first tracing and intermediate views
The account describes selecting processing steps and retaining inspectable intermediate results and parameters. Its automated quality checks are internal.
- NextArchaeologist
Edit, accept or reject suggested features
The archaeologist inspects suggested lines, edits or rejects them and adds annotations in the described workspace.
- NextRA3I documentation export
Export tracings and documentation
The sources describe tracing and overlay views, processing records, motif information and a compiled PDF export.
What the source reports
Seth Chagi describes RA3I profiling a field photograph, selecting processing steps and producing a first tracing with inspectable intermediate results. An archaeologist edits, rejects or accepts suggested lines and annotations. RA3I’s own site describes full-resolution overlays, processing records and PDF export. These are published workflow descriptions, not inspected archaeological project outputs.
Reported technology: RA3I image-processing tools and archaeologist review. Scope and outcomes are source-reported.
What the enterprise environment would need
The following is a proposed project design for an enterprise doing similar work. Published details are marked; the remaining systems, data and scenarios would be agreed for your project.
Proposed practice environment
Review one permitted rock-art documentation bundle
Illustrative project design to scope with your enterprise.
- 1 · Set up the environment
Permitted photographs and frozen specialist references
Use permitted example or synthetic photographs with rights, provenance, image-quality criteria and frozen specialist annotations in an isolated workspace. Supply field or 3D references where needed; missing essential evidence stops the task.
- 2 · Preparation and agent contributions
Agent prepares a reviewable tracing draft
Preserve raw photographs alongside suggested lines, intermediate views, uncertainty and a draft overlay, process record and PDF. Distinguish agent suggestions from human edits; enable no real upload or publication endpoint.
- 3 · Independent checks
A separate archaeologist compares raw evidence
An archaeological specialist separate from the generating agent compares each suggested feature with raw imagery and appropriate field or 3D references. Check rights, provenance and export alignment. Log uncertain contours; internal image-processing checks cannot replace this review.
- 4 · People decide acceptance
Documentation owner accepts an exact bundle
The owner accepts or rejects the exact independently checked rehearsal tracing, overlays, process record, PDF and findings together. Acceptance grants no interpretation, archive deposit or public-release authority.
A reviewable evidence package
A versioned draft tracing and documentation bundle with raw-image references, uncertainty, independent findings and owner disposition.
If a check fails → withhold acceptance.
Missing rights or references, inadequate lighting, uncertain contours, changed imagery, misaligned exports or owner rejection withholds acceptance. Repair the affected bundle or obtain permitted evidence, rerun independent archaeological and provenance checks, then return the exact revised bundle to the owner.
Failure or owner rejection returns to review
- Hold
- → 2 · Revise work
- → 3 · Rerun independent checks
- → 4 · Owner review
Operating boundary. Designed checks, not executed. Permitted example or synthetic rehearsal only; no actual site upload, archive deposit, publication, benchmark result or human-readiness claim. Archaeological and rights owners must confirm reference rules and acceptance criteria.
- Permitted example photographs and archaeological references
- Proposed: a documentation owner supplies approved example or synthetic photographs, rights and provenance records, image-quality criteria and frozen specialist reference annotations. Provide appropriate field or 3D references where interpretation requires them; stop if essential evidence is absent.
- Reviewable tracing and documentation drafts
- Proposed: an agent prepares candidate lines, intermediate views, uncertainty notes and a draft overlay/process-record/PDF bundle in an isolated workspace. Preserve the raw photograph and distinguish generated suggestions from human edits; enable no real upload or publication endpoint.
- Independent archaeological review
- Proposed: an archaeological specialist separate from the generating agent compares every proposed feature with the raw photograph and appropriate field or 3D references, checks provenance and export alignment, and records ambiguous or unsupported lines. Image enhancement cannot establish a feature absent from the evidence.
- Documentation-owner acceptance and repair
- Proposed: the accountable documentation owner accepts or rejects the exact independently checked rehearsal bundle. Rights gaps, inadequate photographs, unresolved ambiguity or changed references block acceptance; repair or obtain permitted evidence, repeat independent specialist checks and return the new version to the owner.
A proposed simulation
An illustrative practice design derived from this work, not an executed assessment.
Proposed: assemble a reviewable draft rock-art documentation bundle from permitted example photographs, preserving raw-image references, intermediate views and unresolved interpretation.
Review gate: The documentation owner accepts only the exact rehearsal bundle whose tracing, raw-photo support, references, rights, uncertainty and exports passed independent archaeological specialist checks. Acceptance grants no archive deposit or publication authority.
Challenge the result: A suggested line follows a lighting artifact rather than a supported feature. Withhold acceptance, repair the tracing and every affected export, repeat independent raw-photo/reference checks and return the exact revised bundle to the documentation owner.
What needs checking
- Source boundary: the author is affiliated with IPT/ITM, a platform partner. RA3I’s own documentation supports the described review/export process, not independent accuracy, privacy, adoption or benchmark validation.
- Proposed: test a supported contour, a lighting artifact, an ambiguous feature, a low-quality photograph, missing image rights and a misaligned full-resolution overlay against separately prepared specialist references. Stop unsupported interpretation. Designed, not executed.
- Proposed: change a photograph, reference annotation or tracing after review. Withdraw acceptance, repair the affected overlay, process record and PDF, rerun independent specialist and provenance checks and return the exact new bundle to the owner. Designed, not executed.
Review environment fidelity, human capability and agent performance separately. The enterprise’s accountable owners decide acceptance of the shared workflow and any operating use.
How an enterprise could build similar work
A proposed build order for an enterprise doing similar work, written by Nuvepro from the published account and our interpretation of the environment and checks above. It is not how RA3I platform described by Seth Chagi built or operates this work.
- Freeze permitted photographs and referencesProposed: agree rights, provenance, image-quality criteria and specialist annotations, including required field or 3D references.
- Prepare reviewable documentation draftsProposed: retain raw imagery, suggested lines, intermediates and uncertainty alongside a draft overlay, process record and PDF.
- Check archaeological support independentlyProposed: a separate specialist checks each feature against raw imagery and appropriate references, plus provenance and export alignment.
- Repair and recheck the whole bundleProposed: stop rights gaps, poor imagery and unsupported lines; repair affected tracings and exports, then repeat independent checks.
- Owner accepts the exact rehearsal bundleProposed: the documentation owner accepts or rejects the independently checked version without archive-deposit or publication authority.
Your systems, data, rules and accountable owners decide the real project. Agree acceptance criteria, escalation owners and operating responsibility with the people accountable for the work before anything runs on live records. Treat the first run as a rehearsal on test records: compare the draft output with an independently prepared answer, and record what the people reviewing it change.
Questions about this example
- What workflow is described?
- Photograph profiling and processing produce a first tracing and intermediate views; an archaeologist edits, rejects or accepts suggestions before documentation export. The platform describes full-resolution overlays, processing records and PDF output.
- Was RA3I independently tested for this example?
- No. The author is affiliated with IPT/ITM, a platform partner, and corroboration is RA3I’s own documentation. No project run, benchmark, privacy audit or linked IEEE paper was inspected.
- What would the separate proposed rehearsal add?
- Permitted example photographs and frozen references, independent archaeological specialist checks against raw imagery and appropriate field or 3D evidence, and exact-version documentation-owner acceptance. These checks have not been executed.
- When must the proposed work stop?
- Missing rights, inadequate photographs or unresolved interpretation block acceptance. Repair or obtain permitted evidence, repeat independent checks and return the revised bundle to the owner; no real upload, archive deposit or publication is authorized.
Shape this around your enterprise
Bring a similar workflow and the people responsible for it. Together, define the records and systems to use, the result your team will build or configure, and how its owners will check it. Nuvepro scopes the simulation, environment and preparation. Agree any existing-agent evaluation, implementation assistance and production responsibilities separately.
Sources and limitations
Curation review: . This is the date this source example was reviewed, separate from source publication dates.
- The public Substack account and primary platform method/output descriptions were read. No real project, tool run, upload, benchmark or privacy audit was inspected; the linked IEEE paper was not retrieved.
- The author’s bio states an IPT/ITM partner affiliation. Platform corroboration is vendor documentation, not independent validation. No numeric gains, customer adoption or delivered enterprise case are established.
- Different described user groups are not separate deployments. Human edit/accept/reject controls described in the sources are not proof that a particular tracing is archaeologically valid.
- The separate proposed rehearsal and its checks are designed and unexecuted. Specialists must confirm domain references and acceptance criteria; a checked draft bundle grants no archaeological interpretation, archive deposit or public-release authority.