Matema Launches OEM Supply of Scheduling Optimization Engine That Mathematically Guarantees Zero Regulatory and Certification Violations for Shift and Security Staffing SaaS — Recalculates in Seconds Without Disrupting Confirmed Shifts During Staff Absences

Matema (headquartered in Aichi Prefecture), a research and development company specializing in mathematical optimization and matching theory, has launched OEM distribution of a 'scheduling optimization engine' designed to be integrated into operational SaaS platforms such as attendance management, shift scheduling, and security dispatch. The engine generates schedules that mathematically guarantee compliance with all constraints—including certification staffing obligations, consecutive work limits, compatibility, and preferred days off—within seconds. When staff absences occur, it recalculates with minimal changes, preserving already confirmed shifts. SaaS providers can now integrate 'AI-powered automatic scheduling'—a key competitive feature—into their platforms at a fraction of the time and cost of in-house development, while offering a 'zero-violation guaranteed' solution that surpasses competitors.

Try the technical demo in your browser

Minimal-change rescheduling upon staff absence—only one cell changes; all other employees' schedules remain untouched

Demo released—see violations of 'AI scheduling that fills in one second' instantly

The publicly available technical demo uses a security staffing scenario (8 officers × 3 sites × 1 week) and allows users to compute the following three steps directly in the browser:

(1) Create a schedule using the sequential 'AI automatic scheduling' method now common in the industry → Fills in within milliseconds, but results in 14 constraint violations, including 5 consecutive workdays, weekly hour overages, and incompatible pairings

(2) Solve the same problem using our engine → Generates a fully compliant schedule, with all automated verification checks marked '✓'

(3) A certified officer calls in sick on Thursday → Recalculate while keeping Monday to Wednesday assignments fixed; only one person's shift is changed, completed in milliseconds. All constraints, including certification staffing rules, are maintained

Schedule generated by the engine and automated verification results—full compliance verified

Background—'Filled' and 'Compliant' Are Different Problems

Since the enforcement of labor reform laws, such as overtime limits for doctors and drivers, shift and staffing operations have shifted from 'filling slots' to 'ensuring compliance with legal, certification, and labor regulations.' In response, SaaS platforms for shift management, security dispatch, and caregiving support have increasingly adopted 'AI-powered automatic generation' features.

However, most of these systems rely on replicating past patterns or sequentially assigning available staff. While these methods can quickly 'fill' a schedule, they offer no guarantee that all constraints—such as consecutive work limits or mandatory certified staffing—are simultaneously satisfied across the entire plan. As a result, field managers often need to manually correct outputs, or schedules are implemented with violations still present. Our engine addresses this by solving the problem as a rigorous optimization task based on constraint satisfaction and integer programming, delivering schedules that are not 'accidentally compliant' but 'verified compliant.'

Measured comparison of violation counts for the same problem (8 officers × 3 sites × 1 week). Breakdown for sequential method: 4 cases of 5+ consecutive workdays, 4 cases of weekly hour overages, 6 instances of incompatible pairs assigned together

Features—Guarantee, Minimal Change, Complex Problem Support

(1) Mathematical guarantee of constraint satisfaction: Simultaneously satisfies certification staffing obligations, 36-agreement limits, work-hour caps, rest intervals between shifts, consecutive work rules, compatibility and pairing conditions, preferred days off, and fairness. Full automated verification results are provided alongside the generated schedule.

(2) Minimal-change rescheduling during absences: The core of staffing operations is not initial creation but reconfiguration. Through re-optimization focused on minimizing changes, the engine adjusts schedules without affecting employees unrelated to the change, preserving confirmed assignments.

(3) Extension to complex problems: Handles integrated challenges such as scheduling combined with visitation routes or shuttle logistics, high-volume daily assignments (dispatch, spot work), and assignments incorporating individual preferences and confirmations (application of matching theory), all within the same framework.

Full rescheduling upon a certified staff member's Thursday absence—Monday to Wednesday assignments fixed, only one person's shift changed

Delivery Model—Phased Adoption from Technical Evaluation (PoC)

· Technical Evaluation / PoC: Use your actual product data and constraints to quickly demonstrate constraint modeling, solvability, and computation time

· OEM Integration: Progress from PoC to API integration or library embedding. We do not sell directly to end users; instead, we enable SaaS providers to offer this as their own branded feature

Request OEM or technical evaluation consultation

Development Background

Over seven years of consulting in logistics, energy, and other operational fields, we repeatedly observed that 'theoretically optimal solutions' often fail to be used in real-world environments where absences and sudden changes occur. The engine's minimal-change rescheduling capability is our response to this real-world challenge. Our mission, reflected in our company name 'Matema' (from 'mathema,' meaning 'that which is to be learned'), is to transform research into systems that operate reliably every day within business operations.

Future Plans

We will deepen collaborations with SaaS providers in security dispatch, workforce/shift management, caregiving/healthcare, and staffing dispatch, while progressively integrating our research on assignment and matching algorithms—presented at international academic conferences—into the engine.