ASSIGNMENT MATCHING
Do not assign cases to people by instinct alone.
Qualifications, experience, regions, availability, continuity, workload, and cost all matter. We build custom systems that evaluate these conditions together and show the best candidates with clear reasons.
- Separate hard rules, preferences, and global objectives
- Show reasons and alternatives
- People review before confirmation
MATCHING BOARDPlan 03
Emergency cooling equipment inspectionM-184 / Kitahama Cold Storage
09:30Control panel replacementM-191 / Central Logistics
13:00Monthly maintenance visitM-204 / Nishimachi Hospital
TodayTHE REAL PROBLEM
Handing work to whoever is free does not create good assignments.
Every case has required qualifications and experience. Every person has strengths, regions, capacity, existing customers, and current load. A single assignment can look fine while the overall plan creates overload and missed opportunities.
Required capabilities cannot be missed
Qualifications, product knowledge, industry experience, language, and approval authority differ by case.
Customer continuity should be protected
Past work, relationships, handover cost, and fit affect assignment quality even when they are hard to quantify.
The capable few become overloaded
When work concentrates on senior people, overtime, dependency, and stalled training follow. The model looks at difficulty and time, not just counts.
Urgent work and absence force replanning
When new work, absences, or priority changes arrive, fixed relationships can be preserved while the remaining cases are reassigned.
We turn the comparison living in a dispatcher or manager's head into explainable calculation and review screens.
MODEL THE DECISION
Design assignment rules in three separate layers.
Separate conditions that must be satisfied, preferences that should be prioritized, and objectives that improve the whole plan so every recommendation can be explained.
model sketch
feasible(person, case) -> score(skill, continuity, response, balance) -> candidate list
The system does not automatically decide people from one score. It separates hard rules, reasons, alternatives, and overall impact for review.
HARD
Conditions that must hold
- Required qualifications and authority
- Availability and capacity
- Allowed region and contract limits
SOFT
Preferences to honor when possible
- Similar case experience
- Customer continuity
- Distance and response time
OBJECTIVE
Metrics to improve globally
- Increase fit scores
- Reduce unassigned cases
- Balance workload
WHERE IT FITS
For work where case requirements and people's strengths must be matched repeatedly.
It fits when there are multiple candidates, many conditions, workload and continuity matter, and changes happen often.
FIELDRepair, maintenance, and inspection
Recommend field technicians from qualifications, product experience, urgency, region, parts, and customer continuity.
BUILDConstruction and site work
Assign leads and support members by trade, required license, skill, site distance, schedule, and crew structure.
SALESEnterprise sales and account ownership
Choose owners and support staff from industry knowledge, product, region, customer size, relationship history, and case volume.
PROJECTProjects and expert staffing
Assign members to projects by technology, experience, period, utilization, role, and team composition.
INTERACTIVE EXAMPLE
Change objective weights and compare candidates with reasons.
This is a simplified explanatory calculation. Real projects include authority, fixed owners, multi-person teams, period planning, revenue, training, and system integration.
This browser demo is not a production optimization engine; it is a way to inspect the decision structure.
Assigned
Required skills met
Average fit
Max load spread
Continuity kept
Unassigned
CASE ASSIGNMENTS
Candidate by case
WORKLOAD
Load by person
UNASSIGNED
Unassigned cases and reasons
RULE DESIGN
Make candidate rules visible as data.
Separate qualifications, capacity, region, continuity, workload, and prohibited attributes so eligibility and ranking do not get mixed.
R001skillRequired capability
Only people who satisfy required capabilities remain candidates.
R002capacityCapacity limit
Projected load must stay within the person's capacity.
R003regionAllowed region
Exclude people outside allowed regions or contract boundaries.
R004continuityCustomer continuity
Evaluate handover cost and prior customer relationship.
R005experienceSimilar experience
Reflect industry, product, and difficulty experience in ranking.
R006responseDistance and response
Evaluate travel or first-response time when it matters.
R007balanceWorkload balance
Avoid concentrating hard work on the same senior people.
R008freezeFreezing and approval
Freeze started or customer-notified assignments when needed.
R009sensitiveProhibited attributes
Do not use age, gender, nationality, health data, or similar attributes.
EXPLAINABLE OUTPUT
Show the material for deciding, not just a person's name.
The screen should show reasons, alternatives, exclusions, and impact on total workload.
Reasons and score breakdown
Show qualification, experience, continuity, distance, and workload scores.
Alternatives and differences
Compare second and third candidates with the same scale.
Reasons for exclusion or no assignment
Classify missing qualifications, capacity excess, region mismatch, conflicts, and other blockers.
Overall impact comparison
Show how manual changes affect remaining unassigned cases, load spread, and continuity.
HUMAN CONTROL
Because people are involved, the automation boundary must be explicit.
This page supports assignment of work to existing staff. It does not design automatic decisions for hiring, termination, promotion, compensation, or other high-impact HR decisions from a single score.
Do not score sensitive attributes
Age, gender, nationality, health information, and similar attributes are excluded from candidate scores.
Require human approval
The system recommends candidates; a manager checks reasons and alternatives before confirmation.
Record changes and exceptions
Manual changes, freezes, exclusions, approvers, time, and reasons are written to audit logs.
Review bias regularly
Track distribution of workload, difficulty, training opportunities, and revenue opportunities.
START WITH YOUR DATA
Check whether matching is possible with a small set of cases and people.
We can prototype candidate calculation from the assignment sheets and rules you already use. An anonymized day or month is enough to begin.
Case and customer list
Due dates, region, difficulty, required capabilities, priority, continuity preference
People and capability list
Qualifications, experience, allowed areas, capacity, limits, preferences
Assignment rules
Required, preferred, excluded, fixed, conflict, and approval rules
Past assignments and results
Owner, time spent, outcome, changes, unassigned cases, and reasons
Validate candidate generation first
Use anonymized case and staff data to compare the current plan with calculated candidates.
- List of hard rules, preferences, and objectives
- Candidates and reasons for each case
- Unassigned and excluded-candidate reasons
- Fit, continuity, and workload comparison
- Proposed production system structure
Check fit for free
If a packaged tool or operations cleanup should come first, we will say so.
FIT CHECK
When custom development fits and when it does not.
We first check whether candidate count, rule complexity, change frequency, and explainability justify a system.
Good fit for consultation
- Each case has different qualification, experience, or region requirements
- Assignment decisions depend on a few managers' tacit knowledge
- Overload on capable people or customer handover is a problem
- Urgent cases and absences require frequent reassignment
- You need reasons and manual-change logs
Clarify rules or use a packaged tool first
- There are few people and fixed ownership works
- Case requirements or capability data are not defined yet
- Simple sequence or region split is sufficient
- Case intake or actual work records should be built first
- The target is hiring, promotion, or other high-impact HR decisions
Next paths
Move from assignment matching to related pages and validation assets.
Shift scheduling
Create rosters satisfying required staffing, qualifications, fairness, and requests.
Visit scheduling
Plan visits with time windows, travel, continuity, and staff limits.
Vehicle routing
Build routes that satisfy capacity, time windows, and vehicle conditions.
Production scheduling
Plan operations across resources, due dates, setup, and downtime.
Demo hub
Compare the five mathematical demos and decide which workflow to validate first.
Case studies
Review how evidence is presented without overstating expected results.
Math Lab
Review the thinking behind mathematical models, validation, and reproducibility.
FAQ
Frequently Asked Questions
Common questions before consultation.
Can we start if skills are not scored yet?
Yes. We can begin from qualifications, past assignments, self-reporting, manager judgment, and case history, then refine the levels gradually.
Can required qualifications and preferred experience be separated?
Yes. Required qualifications decide eligibility, while experience can influence ranking. We keep each condition and its priority visible.
Can one case be assigned to a team?
Yes. We can define roles such as lead, licensed person, and support member, then check whether the whole team satisfies the requirements.
Can customer continuity be prioritized?
Yes. Prior owners, main contacts, handover status, and customer preferences can be modeled and compared against reassignment.
Can the system reassign work when someone is absent?
Yes. Confirmed or started cases can be frozen while only affected unstarted cases are recalculated.
Does the system automatically confirm the assignee?
Usually no. We recommend candidate presentation with human review, correction, and approval. Limited automatic confirmation is possible only for low-risk standardized cases.
Do you use age, gender, or similar attributes?
No. This page assumes support for assigning work to existing staff and excludes sensitive attributes from fit scores.
Is generative AI choosing people?
The core is explicit constraints, mathematical optimization, and search. AI may help extract requirements from text, but eligibility and reasons should remain structured.
Can it connect to our CRM or workforce system?
Yes. API, CSV, database, and file integration can be considered. It can also add only candidate calculation and approval screens to existing systems.
How many cases and people can it handle?
Difficulty depends on teams, periods, fixed assignments, skills, conflicts, objectives, and response time, not just count. We measure performance with realistic sample data.
CHECK YOUR ASSIGNMENT
Your assignment process can become an explainable calculation.
We can look at your current Excel sheet or case list and identify whether the problem fits, which data is needed, and what to prototype first.