Atomic AoL logo Atomic AoL Accreditation & AssuranceOfLearning · a project by Marino&Carli
AACSB Standard 5 · Assurance of Learning

Your accreditation report,
in 30 minutes not 3 weeks.

Atomic AoL is a pipeline of 41 specialised AI agents that turns your grades, surveys and curriculum data into a complete, digitally-signed AACSB Assurance of Learning submission bundle — with a human approving every decision that matters.

0
specialised agents
0
automated tests
0
demo pipeline run
0
per real programme cycle
Born from accreditation research at the Open University Business School.

Three weeks of evidence-chasing, condensed into a coffee break.
The AI does the typing. Your faculty makes the calls.

— the principle behind every one of our 41 agents
The problem

AoL reporting consumes your faculty

Every AACSB cycle, accreditation teams spend weeks pulling grades from spreadsheets, chasing survey results, mapping learning outcomes by hand and writing the same narrative chapters again — or pay £10–50k for external consultants to do it for them.

3+ weeks
typical staff time per AoL report
£10–50k
typical external consultancy cost
30 min
operator time with Atomic AoL
~£0.10
AI cost per pipeline run
The platform

41 agents · 3 layers · 1 signed bundle

Each agent does one job well. Together they execute the whole AACSB Standard 5 workflow.

Layer 1 · Connect

Read your world

Agents ingest the AACSB standard, your qualifications, modules, learning outcomes and assessments — and propose the competency goals that anchor the report.

Layer 2 · Analyse

Find the signal

Grades and indirect measures become cohort aggregates, multi-year trends and benchmark comparisons — with breaches flagged and findings balanced into a defensible narrative arc.

Layer 3 · Narrate

Write & sign

Narrator agents draft the executive summary, evidence paragraphs and gap watchlist, then package everything into a PDF + evidence bundle sealed with a SHA-256 manifest.

Why it's different

Not a chatbot. A workflow that completes.

Human-in-the-loop by design

The AI proposes — your faculty decides. Competency goals and key judgements stop at an approval queue until a named person signs them off. Every approval is recorded.

Audit-ready by construction

Every bundle ships with a cryptographic SHA-256 manifest chaining the PDF to its evidence. A peer-review team can re-derive the hashes and verify nothing was altered.

Your data stays yours

Tenant isolation is enforced server-side on every database row. Data is processed via Anthropic's commercial API, which does not train on your data. EU hosting available.

From data to bundle

Three steps for the operator

1

Upload

Grades CSV, survey results, module list — guided by a web wizard with templates for every file. No IT project required.

2

Approve

Review the proposed competency goals and key findings in the approval queue. Accept, reject or amend — you are the editor.

3

Download

One click runs the full pipeline. Minutes later your signed submission bundle is ready: REPORT.PDF + EVIDENCE.JSONL + MANIFEST.

See it on your own programme

One fixed-price pilot · one full AoL cycle on your data · delivered in two weeks. Verify the quality before any annual commitment.

View pilot pricing → Book a call →