Manual scheduling. Legacy software. XShift AI. 31 rows of what each one actually costs your vet clinic.

The AI Copilot generates the full week in seconds — validating 13 constraints per staffer simultaneously: approved PTO (auto-excluded), the roles each shift requires against the roles each staffer carries, conflicts (Ava and Johnny stay off the same shift), minimum rest between shifts (Johnny gets at least a day between his), max hours per week, max shifts per week, min shifts per week, running OT exposure, availability windows, employee preferences for days and shifts, weekly labor budget per location, reliability of who actually shows up, and pairing requirements where they apply. Manager reviews and publishes in a couple minutes. 250-400 hours of senior-staff time reclaimed every year and routed back to client communication, surgery coordination, and recall outreach — work that actually drives clinic revenue.

Encode Dr. Reyes as a role of his own (one role per surgical DVM you want pair-locked). Then set a staffing rule: "Clinic A, Tuesday 7 AM-1 PM surgery block: require 1 Dr-Reyes role + 1 anesthesia-tech role on the floor." The AI Copilot honors the staffing rule during weekly generation. Any assignment that would leave the surgery block under-staffed for the pair gets blocked at save time. Bumped surgeries from pairing errors: zero.

Autopilot Call-Off filters every eligible tech in under a second — location, role (including secondary roles when multi-role is on), full-day and time-window unavailability, custom shift-assignment rules, schedule conflict + overtime exposure, approved PTO. Then either auto-assigns the most-eligible tech (when the call-off is ≥ the day threshold) or fires in-app pickup messages with one-tap accept to the qualified pool (when it's inside the threshold). Most call-offs cover in 5-15 minutes. Canceled surgery days from coverage gaps: near zero.

Workforce Insights surfaces scheduled hours across every employment classification in one view — W-2 hourly, W-2 salaried, 1099 contractor. The salaried 51-hour scheduled week is visible on the dashboard at the end of week 2, not after the resignation letter lands. Conversation happens early. Cascade avoided.

The Manage tab shows real-time on-floor counts at every hour of the day, with lunch coverage and break status surfaced inline. Front desk sees Dr. Mendez has a free chair AND Tech Brooke is on the floor (not on lunch). Books the sick visit in 30 seconds. Same-day slots fill instead of leaking.

Role-based assignment. A tech is only eligible for DEA-handling shifts if she is assigned the "DEA-eligible" role on her staffer profile. The AI Copilot will not generate a non-eligible assignment. The schedule grid blocks them at save time. Cert expiry itself lives in your HR system — the manager updates the role assignment when the cert renews or lapses, and the schedule responds instantly.

Custom rule on the Autopilot page in natural language: "no shifts within 11 hours of each other for the same staffer." Rule fires at assignment time. AI Copilot will not generate violating shifts during weekly generation. The grid blocks any back-to-back at save time, or shows a confirmation modal the manager can override with a documented reason in a true exception.

On the Autopilot page, you write a custom rule in plain English: "Surgical techs: max 4 shifts/week." The rule fires at assignment time, and the AI Copilot respects the cap during weekly generation. The grid blocks any assignment that would exceed the rule at save time, or shows a confirmation modal the manager can override with a documented reason. Senior-tech burnout from sustained over-scheduling: structurally prevented, not just monitored.

On the Autopilot page, you write a custom rule in plain English: "Maria: 3 shifts/week guaranteed. Lisa: 2 shifts/week." The rule fires at assignment time, and the AI Copilot honors both the floor and the ceiling during weekly generation. Part-time techs stay because the practice honors the agreement they signed for.

Custom rule: "Every DVM-scheduled shift must have a minimum of 2 techs scheduled in the same hours." The Copilot generates with the ratio held. The grid blocks under-ratio days at save time.

On the Autopilot page, you write a custom rule in plain English: "at least 2 front-desk staff scheduled in every 30-minute window between 8 AM and 6 PM at Clinic A." The rule fires at assignment time, and the AI Copilot generates schedules that respect the coverage minimums. Lunch coverage holds. Voicemail almost never fires during clinic hours.

The Autopilot Overtime Scanner flags any assignment that would push a staffer into the 41st hour BEFORE the shift is saved. Manager swaps to a non-OT tech or accepts the OT with eyes open. OT bleed from missed catches: near zero. Across 5 locations: $130K-$200K/year of premium that never gets paid.

Shift trades route through one of three configurable approval modes — the manager picks the mode that fits the org. AUTO_APPROVE: the swap auto-qualifies against every rule and publishes immediately if it passes. MANAGER_APPROVAL: the swap auto-qualifies and then routes to the manager for one-tap approval. CONDITIONAL (Smart Approval): the rules decide automatically for clean cases; edge cases route to the manager. Either way, the swap cannot even be requested if it would violate weekly cap, minimum rest, OT exposure, or role match.

On the Autopilot page, you write a PTO rule in plain English: "auto-deny PTO requests that fall between [date range] at [Clinic A]." Whenever an employee submits a PTO request from their dashboard, the rule fires and auto-denies before the request ever hits the manager's inbox. The employee gets an instant answer plus an email showing the exact rule that fired and why. The manager doesn't spend a single minute approving PTO during your busiest weeks — it's already handled. High-revenue clinic weeks stay staffed.

Multi-location organization out of the box. One organization, many locations, one staffer roster. The AI Copilot sees every tech across every location during weekly generation. Double-booking is blocked at assignment time, no add-on required.

Encode equine-restraint training as a role on the tech profiles — only equine-trained techs carry the role. Then set a staffing rule: "mobile vet truck shifts 8 AM-6 PM require 1 DVM role + 1 equine-restraint role on the truck." The AI Copilot won't generate a route shift without both roles staffed. The grid blocks any reassignment that would break the pair. Day ends on time.

Role-based assignment. Specialty consult shifts are only eligible to be filled by staffers carrying the matching specialty role. The grid blocks non-matching booking at assignment time. Front desk cannot accidentally route a derm case to a non-derm DVM.

Encode grief-trained as a role on the tech profiles — only staffers who have done the training carry the role. Then set a staffing rule for the time windows where euthanasia appointments concentrate: "Clinic A, weekdays 1 PM-6 PM, require 1 grief-trained role on the floor." The Copilot won't generate the block without that role staffed. Front desk only sees grief-trained techs as schedulable for euthanasia bookings at that hour.

Encode dental-trained as a role on the tech profiles. Then set a staffing rule: "Clinic A, dental cleaning shifts 9 AM-2 PM, require 1 DVM role + 1 dental-trained tech role + 1 anesthesia-tech role on the floor." All three locked in by the Copilot at weekly generation. The grid blocks any assignment that would leave any of the three roles short. Dental day stays a dental day.

Announcements broadcast in-app, scoped by location, role, or full org. Manager sees read receipts by staffer. Re-pings unread before training starts. Attendance tracked.

On the Autopilot page, you set ranked PTO rules in plain English. Examples: "AUTO_DENY if the request is within 7 days of the start date." "AUTO_DENY if the request lands on a company-designated holiday." "AUTO_DENY if approval would drop a location below minimum staffing." "AUTO_APPROVE if the request is more than 30 days out and the per-day PTO overlap count is under our limit." Rules sort by priority — the highest-ranked matching rule fires first. You can also pre-block specific date windows (spay-neuter clinic week, December 22-31, the Friday of every dental promo) so PTO during those windows auto-denies without a single manager touch. Every auto-decision sends the employee an email citing the exact rule that fired — no awkward "why did she get it" conversations, no inconsistent yes/no calls across the team. The manager can still manually approve or deny outside the rule set when judgment calls matter. 80-95% of PTO traffic processes itself, the manager owns the exceptions, and employees get same-day answers with a clear reason — instead of waiting 3-5 days for a yes/no from an overwhelmed manager.

Equipment can be modeled as a role on a virtual staffer, OR governed by a custom rule written on the Autopilot page in plain English: "ultrasound bookings: max 1 active assignment per hour." The Autopilot rule fires at assignment time, and the AI Copilot will not generate a conflicting assignment during weekly generation. Equipment time stays clean.

In regular Settings, turn on role-based staffing. Then in Advanced Settings, enable multi-role assignments. On each staffer's profile, set a primary role (e.g., "Anesthesia tech") AND optional secondary roles (e.g., "Dental tech," "Lab tech"). When the AI Copilot generates next week, it fills role-required shifts with PRIMARY-role staffers first. If a shift still needs coverage and no primary-role staffer is available — AND the cross-trained staffer has no other constraint blocking her — the Copilot falls through and assigns her on her secondary role. Cross-training stops being theoretical. Senior techs feel valued, actually get used at the top of their license, and stay.

When a staffer is deactivated on her profile, her existing shifts become open shifts on the schedule. The manager opens the schedule tab and asks the AI Copilot to regenerate. The Copilot fills every open shift in seconds — checking availability, role, max hours, min rest, OT exposure, conflicts, and every other constraint at once — and you don't have to do it yourself. No orphan kennel shifts on Saturday morning. No 6 AM phone calls. The schedule fixes itself.

On the Autopilot page, you write a custom anti-pairing rule in plain English: "Ava and Johnny are never scheduled on the same shift." The rule fires at assignment time, and the AI Copilot will not generate a violating shift during weekly generation. The grid blocks any direct assignment that would pair them. Interpersonal landmines stay defused.

Workforce Insights surfaces scheduled hours per staffer across every classification. On the Reports & Analytics page, the reliability metric (clock-ins divided by assignments) surfaces who consistently shows up. Reliable staffers get the best shifts, the senior assignments, and the performance conversations they have earned — based on real data, not memory.

Workforce Insights rolls up scheduled hours and overtime exposure across every location, week by week, with week-over-week deltas. Owner sees the org-wide OT trend by Tuesday morning. Course-correct in the same week.

Reports & Analytics surfaces days when no staffer carrying a specialty role is scheduled. Manager catches the gap during weekly review and fills it before the calls start coming in.

On the Autopilot page, you write a custom labor-cost rule in plain English: "weekly labor cost over $X at Clinic A: block the assignment." The rule fires at assignment time, not at payroll preview. A shift that would push the day or the week over the cap is blocked at save time, or shows a confirmation modal the manager can override with a written reason.

Schedule published once. In-app notification to every staffer with read-receipt tracking. Manager re-pings unread staffers automatically. Schedule visible by location or org-wide. Schedule confusion shrinks to a rounding error.

Same-day swap requests with rule validation and Manager Approval toggle. Approved swaps publish in real time to the schedule. Off-hours swaps process even when the manager is asleep — if the rules clear and Manager Approval is set to off-hours auto-approve.