Shifts Cover Themselves
The loop spots coverage gaps, ranks eligible workers, offers the open shift and escalates what stays open; at Auto it proposes a fill for approval — you set the autonomy dial.
HOW IT WORKS
How it covers a shift
From spotting an uncovered shift to filling it — using the same eligibility rules and approval chain you already enforce. It starts when there's a gap, not when someone asks.
1. Detect
Checks coverage every 15 minutes. Finds upcoming shifts short of staff, unfilled shifts close to their start, and shifts where a worker called out.
2. Decide
Takes the workers eligible for the shift (right group or location, no overlapping shift, not called out) and ranks them on preferences, availability, reliability, workload, and recency. Picks the strongest match.
3. Act
Offers the shift to the top-ranked workers, notifies managers of the gap, and escalates a shift that is still unfilled close to its start. At Auto, when the top pick is a clear winner, it proposes the fill for a manager's approval. Gap detection, offers and draft schedules are each behind their own switch in the app's automation settings, off by default.
4. Log
Every action the loop takes is written as an autonomous-action record naming the shift, what ran, and whether it completed.
AUTONOMY YOU CONTROL
Three levels of autonomy. You pick.
Start with it off — it takes no action until you say so. Move to approve for a one-tap checkpoint on every action. Let it run on its own when you're ready.
Off — manual only
The loop takes no action on its own. People work coverage through the Scheduling app and chat.
Approve
Each offer or escalation is queued for a manager to approve with one tap. The pending queue is your daily standup.
Auto
Offers to up to three workers, escalations, and coverage notices for shifts more than 4 hours away go out on their own. A wider offer, putting a named worker on a shift, and generating a draft schedule still wait for a manager's approval.
Every shift the autopilot filled shows up with a "Filled by AI" badge
On a shift's page, each assignment the autopilot made carries a small "Filled by AI" badge, and the console's activity feed lists the loop's most recent actions. The pre-AI workflow is preserved; the AI layer is visible, not hidden.
- "Filled by AI" on each assignment the autopilot made, on the shift's page.
- "Done by AI" on the loop's completed actions in the console's recent activity, while AI mode is on; an action that fails reads "AI couldn't complete".
One console — the manager's home for autopilot operations
The AI Shift Manager console is the buyer-facing landing for managers and admins. The unfilled-shift rate sits front and center with a sparkline trend. The "What the AI handled" feed shows the most recent actions. The "AI actions to approve" list surfaces approval-gated fills inline — approve or reject without leaving the page. The console shows the current AI mode; an admin moves it from Approve to Auto on the Automation page once the team trusts what they see.
- Hero metric + trend — unfilled-shift rate with sparkline; the one number that matters.
- "What the AI handled" feed — the most recent offers, coverage notices, escalations, and fills the loop ran, each with a timestamp.
- "AI actions to approve" list — pending approvals approved or rejected inline.
- Autonomy dial — Off, Approve or Auto, set in the app's settings; the console shows the current mode.
Why Shift Coverage Goes Sideways
AI Shift Manager attacks the specific failures that make schedules feel chaotic to employees and managers — without changing the rules and approvals the Scheduling app already enforces.
Employees Forget When They Work
"Do I work Saturday?" is the most-asked question in any shift-based business. Without a fast lookup, the employee texts a coworker, checks the wall printout, or just doesn't show up. A 5-second answer becomes a missed shift.
Open Shifts Sit Unclaimed Until A Manager Begs
Someone called out. The marketplace has the open shift. But the employees who could pick it up don't know it exists — and the manager has to text the team one by one. Hours pass; coverage gaps widen; the shift gets filled at the last minute or not at all.
Swap Requests Take Three Apps And A Manager Reply
The employee wants to trade Wednesday for Friday. They text a coworker. The coworker agrees. They send the manager an email. The manager replies. Then someone has to actually go into the Scheduling app and make the change. Hours of friction for a 5-minute trade.
Availability Is Set Once And Never Updated
The employee set their availability at hire. School schedule changed. Side gig started. Family commitment. Nobody updates the system because the availability UI is buried four screens deep. So they keep getting scheduled into shifts they can't work.
Overtime Sneaks Up On Whoever's Closest To The OT Cap
The employee picks up two open shifts on the marketplace, neither of which surfaces that they're now at 38 hours with a 40-hour soft cap. Payroll cutoff hits, the OT lands, and the manager owes finance an explanation. Eligibility checks have to happen before claim, not after.
Multilingual Frontlines Get English-Only Shift Notifications
The schedule posts at 4pm Friday and the email goes out in English. A third of the crew reads Spanish or Vietnamese first. They open the message Monday morning and miss the Saturday change — and the manager spends the morning fielding "I didn't know I was on" texts.
AI Shift Manager At A Glance
AI Scheduling
Real-time shift coverage. Built for the frontline. Multilingual.
Inside AI Shift Manager — The Actual Capabilities
Every block below maps to a real tool the agent uses against your scheduling data. Read tools surface schedules, availability, and the open-shifts marketplace. Every write — employee-side or manager-side — requires explicit confirmation before it runs.
Your Shifts — This Week, Next Week, Any Week
The most common scheduling question, answered without opening the app. The agent surfaces your assigned shifts with start time, end time, role, location, coworkers, and total scheduled hours — all in one ask.
view_schedule— your OWN assigned shifts. The single most common scheduling query.view_shifts— details for a specific shift including role, location, and coworkers.view_team_schedule— the published shifts of the people a manager oversees, such as their direct reports and the scheduling teams they run; for an employee, the roster their Team Calendar shows.- Permission-aware — employees see their own shifts and their Team Calendar's roster; managers see the people they oversee; nobody sees what their role doesn't allow.
Set Availability — Without Burying It In A Settings Tab
Updates to availability used to require digging through four screens of preferences. The agent makes it a one-prompt update — "make me available Saturday mornings" — with the full grid surfaced for confirmation before any change.
view_availability— your current day/time blocks and preferences (preferred days off, weekly target hours, willingness to pick up extras).set_availability— risky write. Updates the user's availability preferences. The system handles confirmation.check_shift_availability— manager-side; checks who on the team is unscheduled for a date/time.- Targeted updates — set_availability only updates the keys explicitly mentioned. Unspecified preferences stay as-is.
Open Shifts Marketplace — Claim Or Drop
The agent surfaces open shifts you're eligible to claim — filtered by role, location, skills, and conflicts — and gates the actual claim or drop behind confirmation. Eligibility is enforced by the scheduling rules already configured; the agent doesn't bypass them.
view_open_shifts— UNASSIGNED shifts in the marketplace, filtered to those the user is eligible to claim.claim_open_shift— risky write. Requires confirmation. Eligibility (location or group, tier, days off, required skills) and conflict checks enforced before the claim is processed.drop_shift— risky write. Requires confirmation. Posts the shift to the marketplace immediately; no manager approval is needed to post it.request_shift_swap— risky write. Requires confirmation. Lists the shift for trade only and notifies the coworker you name; the trade completes when you accept a coworker's offer of one of their shifts.
Swap, Drop, And Coverage Visibility For Managers
Three flows that traditionally took apps, texts, and emails — now started from chat, each behind an explicit confirmation. A trade is settled between the two employees; a pickup follows the company's claim approval setting. Managers see coverage status.
request_shift_swap— list your shift for trade and notify a specific coworker; they answer by offering one of their own shifts in Shift Marketplace, and the trade is done when you accept.drop_shift— list a shift on the marketplace for coworkers to pick up; you stay assigned and responsible for it until someone claims it and that claim is approved.check_shift_availability(manager view) — surface who could pick up an open or dropped shift.- AI request log on every action — read or write, every tool call is logged with the requesting user and the tool (AI request detail is kept 30 days).
Outcomes Teams Can Measure
The agent is built to compress shift-coverage time, deflect routine "when do I work?" questions, and surface open shifts to people who can actually claim them. Measure against your pre-agent baseline.
- Time to fill an open shift — minutes from a shift hitting the marketplace to it being claimed.
- Schedule-question deflection — manager and supervisor interruptions for "when do I work?" / "is this my shift?" deflected to the agent.
- Open-shift visibility — share of marketplace shifts surfaced to eligible employees within an hour of posting.
- No-show / late-call rate — leading indicator that schedule clarity is improving.
- Swap-and-drop cycle time — hours from a shift being listed for swap or drop to a coworker taking it.
Every Write Confirmed, Eligibility Always Enforced
AI Shift Manager declares 35 tools. 30 are reachable in chat — 12 read-only (schedules, shift details, availability views, open shifts, team schedule, pending shift claims, unfilled shifts, absence reports, coverage gaps, demand forecast) and 18 writes, every one of which requires an explicit confirmation before it runs. The remaining 5 are autonomous-only: they never appear in chat and are authorized by the autonomy settings a manager configures, not by a prompt. Every claim is checked against the same eligibility rules (role, location, skills, conflicts) the Scheduling app enforces.
- 18 confirmation-required writes in chat — employee-side
claim_open_shift,drop_shift,request_shift_swap,set_availability,report_absence,tool_withdraw_absence_report,tool_update_absence_report,tool_withdraw_shift_claim; manager/admin-sidetool_create_shift,tool_assign_user_to_shift,tool_unassign_user_from_shift,tool_publish_schedule,tool_approve_schedule,tool_approve_shift_claim,tool_reject_shift_claim,tool_acknowledge_absence_report,tool_resolve_absence_report,tool_release_absence_shift. None runs without the user confirming. - 5 autonomous-only tools, never chat-reachable —
tool_auto_fill_shift,tool_offer_shift,tool_escalate_unfilled_shift,tool_generate_schedule,tool_notify_coverage_gap. These run from the loop under the autonomy level you set, not from a chat prompt, and are excluded from the schemas the model is shown. Auto-fill and schedule generation always wait for a manager's approval. - Eligibility filters always applied — role match, location, required skills, and conflicts checked before a claim.
- Approval routing respected — a pickup follows the company's claim approval setting, a trade completes only when you accept a coworker's offered shift, and a dropped shift stays yours until a coworker claims it and that claim is approved.
- AI request log on every action — read or write, every tool call is logged with the requesting user and the tool (AI request detail is kept 30 days).
WHAT TEAMS TRY INSTEAD
The four alternatives — and why none of them enforce the eligibility and approval chain the schedule already has
Most teams have piloted scheduling AI. The honest gap is that a chatbot that claims open shifts is dangerous unless eligibility and approvals are enforced the same way the schedule itself enforces them.
Pasting schedules into ChatGPT, Claude, or Copilot
General-purpose AI helping a worker think about an open shift
- Eligibility (role match, location, skills, conflicts) checked before any claim — not after the fact
- Approval routing honored — pickups follow the company's claim approval setting, a trade needs both employees to agree, and a dropped shift stays the employee's until a claim on it is approved
- Permission-aware and logged per tool call, so schedule and personnel data stay under your access rules — AI requests go to the platform's configured model providers
Deputy AI / When I Work AI / 7shifts AI / Connecteam AI
Vendor-trapped scheduling AI inside a separate workforce tool
- Reads from the same Scheduling app records the manager already built — not a synced shadow copy
- Lives where employees already do their full workday — no separate scheduling app sign-in for a swap
- Available to crews who never had a separate scheduling vendor seat in the first place
Custom shift-swap forms and group chats
A manager's spreadsheet patch that the next reorg breaks
- Eligibility filters always applied — group chats let ineligible workers volunteer for shifts they cannot legally take
- Confirmation-gated writes — every claim, drop and trade is confirmed and recorded in the app, not agreed in a chat thread
- Every tool call logged with the requesting user and the tool — AI request detail is kept 30 days
The manual fallback — workers text the manager
A swap process that loses to the manager being off-shift
- Open shifts surface in the worker's own app — no waiting for the manager to broadcast
- Set-availability happens in chat with confirmation, not a phone call after the schedule posted
- Claims waiting for approval reach the manager on the same shift, not a separate text thread
PLATFORM ADVANTAGE
AI Shift Manager inherits everything the app already enforces
A standalone scheduling AI has to plumb eligibility, approval chains, and request logging. AI Shift Manager gets all of it for free.
Confirmation-gated writes
18 chat-reachable writes — claim_open_shift, drop_shift, request_shift_swap, set_availability, report_absence, the withdraw and update of your own report, and the manager-side schedule and absence-report tools — each requires explicit confirmation. The model proposes; the person commits.
Eligibility filters always applied
Role match, location, required skills, and conflicts checked before any claim — the agent enforces what the schedule enforces.
Approval routing respected
Pickups follow the company's claim approval setting, a trade completes only when both employees agree, and a dropped shift stays the employee's until a coworker's claim on it is approved. No shortcut around the rules the app already runs.
Marketplace integration
Open shifts and marketplace listings visible in one prompt — the same data the Marketplace app shows, with the same eligibility gates.
Cross-app data plane
Reads schedule, availability, and skill records the platform already maintains — no parallel eligibility store.
AI request log on every action
Every read and write is logged with the requesting user and the tool (AI request detail is kept 30 days). Each action the loop takes is also written as its own autonomous-action record.
INDUSTRY FIT
Industries where shift coverage decides the day
AI Shift Manager earns its keep where coverage is the binding constraint and eligibility rules are real.
Retail
Associates claim open shifts and set availability between transactions — managers approve from one prompt instead of three texts.
Healthcare
Required skills checked before a claim, and claims wait for a manager when your approval setting says so — so a shift needing a skill your nurse lacks never reaches them.
Manufacturing
Plant-floor claims respect skill match and conflicts — eligibility enforced before anything goes to approval.
Hospitality
Property staff trade shifts with each other and pick up open ones — GMs approve pickups from the same app they use for recognition and shifts.
Logistics & Warehousing
Yard and dock teams pick up open shifts with eligibility checks for location or group, conflicts, and any skill the shift requires.
Public Sector
Open-shift claims follow the approval workflow the schedule already runs, and a trade completes only when both employees agree.
WHY MANGOAPPS WINS
An embedded scheduling agent beats a horizontal AI, a Deputy-class vendor, or a manager spreadsheet on every axis
The argument operations, HR, finance, and frontline supervisors all share — and the one Deputy or 7shifts structurally cannot answer.
Cheaper than the alternatives
No Deputy AI tier, no When I Work add-on, no 7shifts subscription, no engineering team building a custom shift-swap form.
More secure
Confirmation-gated writes, eligibility enforced before claim, every tool call logged. AI requests go to the platform's configured model providers, like every MangoApps agent.
Easier to deploy
Already deployed if Scheduling is on. Agent picks up the active eligibility rules and approval chain the same day.
Easier to use
Frontline workers claim and swap from the same mobile app they already opened for shifts — no separate scheduling tool to learn.
Easier to manage
Rule edits in the app are immediately visible to the agent — no re-training, no parallel rule store, no drift.
Easier to extend
Shares the agentic-tool framework with every other MangoApps agent. New scheduling tools (a new availability shape, a new swap pattern) ship as tools.
AI is actually better
A horizontal AI can suggest a swap. AI Shift Manager enforces eligibility, routes through the existing approval chain, and commits only after explicit confirmation — every tool call logged.
At our head office within HR, marketing, and finance, we use ChatGPT quite a bit, and MangoApps AI Helper is a useful tool for us. As I go around to the stores, I’ve been showing them how the AI Intranet Helper can find things and answer plain language questions about our resources, and they’re quite excited about that.
Alison van Zijl
Head of Human Resources
Full House
Customer Success
MangoApps Customers
Organizations that trust MangoApps to run every workflow across their workforce.
Video Case Studies
Customer Case Studies
Customer Case Studies
Customer Case Studies
Customer Case Studies
Customer Case Studies
Frequently Asked Questions About AI Shift Manager
35 tools across the shift lifecycle. 30 of them are reachable in chat. For employees: view your own assigned schedule, view shift details (role, location, coworkers), view and set your availability, view open shifts in the marketplace, claim an open shift, request a swap with a coworker, drop a shift back to the marketplace, withdraw a pickup request you made, and call out of or report running late for one of your own shifts, then list, update or withdraw your own report. For managers and admins: list schedule records (draft / pending approval / published), view the team schedule, check who is free for a date and time, view the demand forecast, check coverage gaps, list shifts with nobody assigned, list the shift claims waiting on you, create a shift, assign or unassign an employee on a shift, publish a draft schedule, approve a schedule, approve or turn down an employee's shift-claim request, and handle absence reports: list them, acknowledge or resolve one, and release a call-out's shift to the marketplace. The other 5 tools are autonomous-only — the loop uses them, chat never does.
view_open_shifts filters to shifts the employee is eligible for — role match, location, skills, and conflicts. Eligibility is enforced by the scheduling rules already configured in the Scheduling app; the agent doesn't bypass them. If an employee isn't eligible for a shift, it doesn't appear in the marketplace view.
No. Every one of the 18 chat-reachable write tools — claim_open_shift, drop_shift, request_shift_swap, set_availability, report_absence, tool_withdraw_absence_report, tool_update_absence_report, tool_withdraw_shift_claim, and the manager-side tool_create_shift, tool_assign_user_to_shift, tool_unassign_user_from_shift, tool_publish_schedule, tool_approve_schedule, tool_approve_shift_claim, tool_reject_shift_claim, tool_acknowledge_absence_report, tool_resolve_absence_report, tool_release_absence_shift — is declared confirmation-required. The agent shows the parsed action (shift, time, role, coverage impact, who it routes to) and waits for the user to confirm before running anything that changes a schedule record. The 5 autonomous-only tools are the exception by design: they are never offered in chat, and they run under the autonomy level a manager sets on the loop rather than under a chat confirmation.
request_shift_swap lists your shift in Shift Marketplace for trade only and notifies the coworker you name. An eligible coworker — the one you named or another — answers by offering one of their own shifts, and the trade happens when you accept that offer. No manager approves a trade; the company's claim approval setting applies to open-shift pickups.
Time to fill an open shift, schedule-question deflection, open-shift visibility (share surfaced to eligible employees promptly), no-show / late-call rate, and swap-and-drop cycle time. Compare against your pre-agent baseline.
Let's Talk
Every team. Every employee. Every workflow. One AI platform — built for the frontline. Since 2008, trusted by 2 million+ users with an NPS of 78.
Why Choose Us?
- Frontline AI: Governed AI for every employee and workflow.
- Top Security: HITRUST, ISO & SOC 2 certified.
- Exceptional UX: Delightful on mobile and desktop.
- Proven Results: 98% customer retention rate.
Rolled out to every employee at AutoZone (125,000), PetSmart (50,000+), A.S. Watson and Raley's (20,000) — and at larger retailers we are not permitted to name.
Prefer to explore first? Ask AI about AI Shift Manager →