AAtlas
Start learning→

academy

The operator's
curriculum.

Five tracks. 44 units. 8 hours of focused, prompt-first instruction designed for owners, ops leads, and the team you actually have.

Begin track 01 →Open the playground
01

Foundations

6 units · 42 min

→
  1. 01.01What Claude actually does (and doesn't)~6m
  2. 01.02The four operator habits~8m
  3. 01.03Risk, trust, and where to start~10m
  4. 01.04Reading model outputs critically~12m
  5. 01.05Workflow vs. one-off prompting~6m
  6. 01.06Your first reclaim: pick a Tuesday task~8m
02

Prompting for business

9 units · 78 min

→
  1. 02.01Briefs that produce shippable work~6m
  2. 02.02Voice docs and brand consistency~8m
  3. 02.03Role framing & persona prompts~10m
  4. 02.04Structured outputs (JSON, tables)~12m
  5. 02.05Few-shot patterns from your archive~6m
  6. 02.06Evals: how to know it's good~8m
  7. 02.07Long-context: feeding the right things~10m
  8. 02.08Avoiding hallucination in numbers~12m
  9. 02.09Prompt versioning for teams~6m
03

Workflows by department

14 units · 165 min

→
  1. 03.01Finance: AR autopilot~6m
  2. 03.02Finance: weekly cash briefing~8m
  3. 03.03Marketing: brief → assets~10m
  4. 03.04Marketing: post-mortem reports~12m
  5. 03.05Sales: post-call follow-ups~6m
  6. 03.06Sales: CRM hygiene loops~8m
  7. 03.07Support: triage & draft~10m
  8. 03.08Support: macro generation~12m
  9. 03.09Operations: SOPs from Looms~6m
  10. 03.10Operations: contract first pass~8m
  11. 03.11HR: hire-to-productive plans~10m
  12. 03.12HR: policy Q&A bot~12m
  13. 03.13Ecommerce: listing factory~6m
  14. 03.14Ecommerce: review synthesis~8m
04

Integrations & guardrails

8 units · 96 min

→
  1. 04.01QuickBooks (read-only patterns)~6m
  2. 04.02HubSpot (deal hygiene loops)~8m
  3. 04.03Google Workspace (Drive + Gmail)~10m
  4. 04.04Microsoft 365~12m
  5. 04.05Slack (the approval queue)~6m
  6. 04.06DocuSign (contract redlines)~8m
  7. 04.07Notion (your second brain)~10m
  8. 04.08Permission scoping & audit logs~12m
05

Agentic workflows

7 units · 88 min

→
  1. 05.01What 'agentic' actually means~6m
  2. 05.02Multi-step runs with checkpoints~8m
  3. 05.03Tool use: when to grant write access~10m
  4. 05.04Failure modes & rollback patterns~12m
  5. 05.05Cost & latency budgets~6m
  6. 05.06Human-in-the-loop UI patterns~8m
  7. 05.07Your first multi-step workflow~10m

AI maturity score

A team-level diagnostic showing where Claude can help most — and where to wait.

Prompt labs

Hands-on exercises with real datasets. You leave with prompts that work in your business.

Verifiable cert

Earn an Atlas certification you can put in your bio. Re-issued as the model evolves.