Map data and access requirements before building an automation
For a selected business process, identify the records the assistant needs, where they live, and which roles may read or change them. Turn the gaps into a concrete integration brief.
How this works
For a selected business process, identify the records the assistant needs, where they live, and which roles may read or change them. Turn the gaps into a concrete integration brief.
Map the exact records and actions required by one proposed automation before choosing integrations. Distinguish where data exists from whether the intended actor may read or change it, and turn each gap into a concrete preparation task.
Before you start
Defined operating scope
A chosen automation process needs a concrete map of records, systems, relationships, and allowed actions.
Required source material
Obtain Selected process description; System and data inventory; Current access rules. If connectors are unavailable, use authorized files with the same fields.
Decision and exception rules
The same entity has conflicting sources. Ask the business owner to select the authoritative field source.
Prepare the inputs
- process_scope
- Required
Trigger, desired output and actors
Bound the workflow before inventorying systems.
- data_inventory
- Required
Entities, systems and record keys
Identify actual source owners.
- access_rules
- Required
Role-to-action permissions
Separate read, write and execute authority.
- integration_constraints
- Required
Available connectors and export paths
Do not assume a proposed API exists.
Implementation, step by step
Establish the working record
- Confirm the selected records, effective versions and responsible owner.
- Check process_scope, data_inventory, access_rules, integration_constraints before beginning. Record the source version and the selected scope; an unavailable value stays unknown rather than becoming an inferred fact.
Checkpoint: The input record identifies the selected process_scope and any missing required fields.
List the entities and records required by the process
- List the records needed for each workflow step, including identity keys and the minimum fields required.
- Verify every required field has a source, owner, key and access status.
Checkpoint: A saved intermediate result for: List the entities and records required by the process.
Map each item to its source system and relationships
- Map entities to authoritative systems and relationships, noting duplicated or conflicting sources.
- Walk through one sample record across systems to test joins and reveal missing identifiers.
Checkpoint: A saved intermediate result for: Map each item to its source system and relationships.
Record the permitted read, write, and execution actions for each role
- Record allowed read, write and execution actions for each actor, including any approval boundary and unavailable access.
- Verify every required field has a source, owner, key and access status.
Checkpoint: A saved intermediate result for: Record the permitted read, write, and execution actions for each role.
Identify missing data, unclear ownership, and integrations that need preparation
- Prepare an integration brief with missing data, unresolved ownership and feasible connector or file-based pilot paths.
- Walk through one sample record across systems to test joins and reveal missing identifiers.
Checkpoint: A saved intermediate result for: Identify missing data, unclear ownership, and integrations that need preparation.
Package the result and unresolved work
- System owners confirm the sources and access rules. Describing an integration does not grant access or show that the data is complete.
- A map describes requirements; it does not grant access or prove completeness.
Checkpoint: A process-specific data map and integration requirements brief. Unresolved items remain visible in the review record.
Decisions along the way
The same entity has conflicting sources.
Ask the business owner to select the authoritative field source.
A required write action is not permitted.
Design a reviewable export or manual handoff instead of implying the integration can write.
The proposed connector is unavailable.
Use a supported file exchange for a pilot or record an implementation gap.
What you will produce
automation-pilot-data-access-map-result.json
JSONA process-specific data map and integration requirements brief. Label proposed and completed states separately.
Include: process_step · entity · required_fields · source_system · record_key · allowed_actions · owner · preparation_gap · status · evidence_reference · review_note
automation-pilot-data-access-map-review.md
MarkdownKeep the evidence and unresolved decisions needed to apply this result, including exceptions involving integration_constraints.
Include: source_reference · question · owner · next_action
Check the result
- Verify every required field has a source, owner, key and access status.
- Walk through one sample record across systems to test joins and reveal missing identifiers.
- Inspect process_scope and data_inventory in both the original and output; confirm that record identity and interpretation survived the transformation.
- System owners confirm the sources and access rules. Describing an integration does not grant access or show that the data is complete.
When something goes wrong
The same entity has conflicting sources.
Ask the business owner to select the authoritative field source. Retain the affected evidence and recovery status.
A required write action is not permitted.
Design a reviewable export or manual handoff instead of implying the integration can write. Retain the affected evidence and recovery status.
The proposed connector is unavailable.
Use a supported file exchange for a pilot or record an implementation gap. Retain the affected evidence and recovery status.
A worked example
SYNTHETIC EXAMPLE · NOT A REPORTED BUSINESS RESULT
Request
Map data and access requirements before building an automation for the following bounded example; prepare the result for review.
Inputs
Synthetic example: Workflow needs customer IDs from CRM and paid invoices from accounting. CRM permits read access; accounting allows CSV export but no API write.
Expected result
Synthetic expected result: Map customer_id and invoice_customer_id with a join-validation task. Propose read-only CRM plus accounting CSV input; no payment or accounting write permission is implied.
Run a small pilot
Start small
Trial a representative batch containing the ordinary case and the exception demonstrated above.
Record the baseline
Record current preparation time, review effort and correction types for the same records.
Agree success criteria
Agree acceptable review and error limits; measure unresolved data and access prerequisites; system-owner corrections to the integration brief.
Stop or revise
The proposed connector is unavailable. Pause affected records and apply this recovery: Use a supported file exchange for a pilot or record an implementation gap.
Keep the scope clear
- A map describes requirements; it does not grant access or prove completeness.
- Limit the inventory to the selected process rather than conducting an unrelated organization-wide audit.
Check before using
System owners confirm the sources and access rules. Describing an integration does not grant access or show that the data is complete.
Business activities
- Operate and improve AI workflows
FROM A RECIPE TO YOUR WORKFLOW
Bring the company context together.
This workflow needs reliable inputs before it needs a better prompt. Moat8 Vault brings working information into a readable archive for the AI tools you use.
For this recipe: Selected process description · System and data inventory · Current access rules
Start with one recurring task, agree which sources to use, and review the first result with your team.
Discuss this workflow