Skip to main content
WHISKERMATCH
Workflows

Start with one workflow.

WhiskerMatch is not asking your shelter or rescue to replace everything. Choose the handoff that keeps costing time, make its sources and owners visible, and expand only if the team wants to.

Browse workflowsPick one workflow for a pilot
One operating structure

Different pain. The same trust loop.

Every workflow keeps preparation separate from judgment: source attached, AI preparation visible, human reviewer named, next owner assigned, and public changes gated.

Intake cleanup

A busy intake desk should not have to type the same fact three times just to keep it from disappearing.

Today's pain
Facts arrive on paper, in exports, and in free-text notes. Staff retype them and still cannot tell what is unknown.
What changes
One structured intake view keeps source, owner, completeness, and unknown fields visible.
AI can prepare
Structure notes, flag missing fields, and identify likely duplicates for review.
Humans decide
Confirm facts, resolve duplicates, and decide what still needs assessment.
Pilot version
Ten to twenty records with one intake owner and a short required-field list.
Success looks like
Less retyping and fewer unknowns silently becoming assumptions.
Never automate
Medical judgment, behavior judgment, or invented values for unknown fields.

Foster update routing

WhiskerMatch does not replace the foster who knows the animal. It keeps their context from getting lost.

Today's pain
Useful context lands in texts and group chats, then depends on one person remembering where it belongs.
What changes
Updates return to the animal record with a source, owner, and explicit next action.
AI can prepare
Summarize the update, separate observations from requests, and flag missing context.
Humans decide
Whether the update changes care, behavior context, public wording, or follow-up.
Pilot version
One foster group, one routing owner, and one weekly review cadence.
Success looks like
Foster updates stop requiring a search through old messages.
Never automate
Behavior conclusions, medical conclusions, or public profile changes.

Applicant review lane

The point is not to score people faster. It is to keep staff from rebuilding the same context before every reply.

Today's pain
Applications, documents, and questions sit in separate inbox threads without a shared state.
What changes
Each applicant has a visible review stage attached to the animal record.
AI can prepare
Organize documents, summarize the timeline, and draft routine replies.
Humans decide
What information matters, what question comes next, and every applicant outcome.
Pilot version
Five to ten applications with one adoption coordinator and explicit stages.
Success looks like
Fewer ownerless applications and fewer repeated questions.
Never automate
Applicant ranking, rejection, placement decisions, or policy exceptions.

Public profile review gate

Unknown should stay unknown until someone reviews it. A faster profile is not worth a riskier one.

Today's pain
A public profile can lag behind current foster or staff context, or change without a recorded reviewer.
What changes
Drafts stay internal until a named human approves, holds, or requests changes.
AI can prepare
Draft clear public wording from attached, reviewed source material.
Humans decide
What is accurate, safe, useful, and ready to publish.
Pilot version
Three to five profiles with one reviewer and explicit hold reasons.
Success looks like
Correct context reaches the public only after review.
Never automate
Publishing, medical claims, behavior claims, or removal of uncertainty.

Follow-up ownership

The work should not disappear just because the animal left the building.

Today's pain
The next check-in belongs to whoever remembers it, especially after the visible placement work is done.
What changes
Every follow-up keeps an owner, due date, source context, and completion event.
AI can prepare
Draft reminders, organize due work, and flag tasks without owners.
Humans decide
Who follows up, when to escalate, and what action a response requires.
Pilot version
One follow-up sequence for two weeks with a single board owner.
Success looks like
Fewer missed check-ins and fewer tasks with no owner.
Never automate
Escalation decisions, welfare judgment, or outreach that has not been approved.

Director posture board

Visibility should reduce the need to interrupt five people just to learn what is stuck.

Today's pain
Leaders get counts but not the reasons work is blocked or who owns the next move.
What changes
A bounded posture board shows blockers, owners, due work, and review states.
AI can prepare
Summarize status and surface exceptions that match shelter-defined rules.
Humans decide
Priorities, staffing changes, escalation, and policy.
Pilot version
One posture view for a small workflow and one operating meeting.
Success looks like
Leaders can ask a better question without rebuilding the board.
Never automate
Priority decisions, staffing decisions, or placement outcomes.

Volunteer reviewer queue

Volunteers should be able to help without guessing which version the team means.

Today's pain
Volunteers cannot reliably see which draft is current, what source changed, or what staff need reviewed.
What changes
A bounded queue presents the current draft, source material, and allowed review actions.
AI can prepare
Assemble context, highlight changed fields, and draft a concise review summary.
Humans decide
Approve, hold, request changes, or send the item back to staff.
Pilot version
One queue, one reviewer group, and a narrow set of allowed actions.
Success looks like
Reviewers spend time reviewing instead of locating context.
Never automate
Reviewer identity, approval, rejection, or policy interpretation.

Audit and source trail

The goal is not surveillance. It is making the handoff understandable when someone new has to pick it up.

Today's pain
Teams know something changed but cannot quickly see who changed it, why, or from which source.
What changes
Each material change records the actor, source, timing, and review decision.
AI can prepare
Organize events and flag actions missing a source or reviewer.
Humans decide
Whether the source supports the change and whether corrective action is needed.
Pilot version
Track source and review events for the first selected workflow.
Success looks like
A staff member can understand why the record changed.
Never automate
Attribution, approval, deletion of history, or retroactive justification.
Founder-supported pilot.

Pick one workflow for a pilot.

A busy shelter should not need a perfect process to start. Bring the workflow as it really operates, including the awkward parts.

Talk through a pilotRead the pilot brief