Case Study
A global lottery operator cut field-service SLA penalty risk by 20%
25+ states, six work types, and a skill-matched technician workforce, now run by Locus, the agentic decision layer for field service, making scheduling, dispatch, and routing decisions autonomously.
Segment & Geography
- Industry: Field services. Lottery systems installation, maintenance, and repair.
- Region: United States.
Objectives
- Automate dispatch across every state, each with its own contracts, labor laws, and SLA terms.
- Cut SLA penalty risk by covering the tightest response windows first.
- Match six case types to a complex, skill-differentiated roster from one planning model.
Locus Solutions Implemented
- Dispatch Agent: autonomous case-to-technician assignment, matching each case to a qualified technician within each state's contract terms.
- Capacity Agent: technician rosters, skill sets, schedule types, staffing models, and standby time planned across zones and shifts.
- Orchestrator Agent: coordinates dispatch and capacity decisions so assignment and roster planning act as one decision layer.
- Legacy system integration: custom case priorities flow from the client's system of record into every assignment decision.
- Governance layer: every autonomous assignment logged for traceability, with human override at every step.
Impact
states on one autonomous dispatch engine
lower SLA penalty risk
lower fuel spend
Client Overview
The company is a global lottery operator and technology provider. Across the United States, its field-service operation installs, maintains, converts, and repairs lottery machines for state lotteries in 25+ states. No two technicians carry the same skill set, and rosters shift through the day.
The work runs under separate state lottery contracts, each with its own labor laws, revenue terms, and service rules. Some carry response windows measured in hours, with penalties accruing while a machine is down. As the footprint grew, manual dispatch hit its ceiling: one dispatch model had to hold all of it, and no amount of planner effort could.
The provider brought in Locus, the agentic decision layer for field service, to make dispatch and scheduling decisions autonomously across its operations.
Business Challenges
- Every state played by different rules. Contracts, labor laws, and revenue terms differ from state to state. Some states carry one-hour SLAs with liquidated damages of $100+ per hour for every machine left unrepaired. A single dispatch approach could not satisfy all of them at once, and manual tuning for each jurisdiction did not scale.
- Six kinds of work, and no two technicians alike. Installation, preventive maintenance, conversion, new site setup, removal, and service each require different skills, so not every technician can handle every case. Add differing on-site time per case type, and every assignment becomes a three-way match of case, skills, and location that manual dispatch had to solve over and over, all day.
- A schedule too complex to build manually. Different zones, schedule types, staffing models, standby time, and technicians moving on and off shift through the day. Even when planners got the roster, the rules, and the map aligned, the day moved: urgency, traffic, weather, and case status shift constantly against the custom priority the legacy system sets for each case. A schedule built at the start of a shift was stale within the hour.
Solutions Implemented
Locus runs as the agentic decision layer for field service across every state the provider serves. The provider's existing systems stay the system of record, holding case priorities and contract terms; Locus runs as the system of execution, turning those inputs into governed dispatch and scheduling decisions.
Autonomous dispatch that holds every state's rules at once. Locus models each state's contract terms, labor laws, SLA windows, zones, and required skills as situationally aware workflows: live constraints drawn from the 250+ real-world rules the platform can hold per computation. Cases route to eligible technicians automatically, with each jurisdiction's rules applied on every assignment. The engine protects the tightest response windows first.
One assignment engine for six kinds of work. The Dispatch Agent assigns every case type through one engine, matching each case to a technician qualified for it while balancing priority, time, and distance across the workforce. Trade-offs between urgent repairs and scheduled work are computed, case by case.
Scheduling that re-plans itself as the day moves. The Capacity Agent holds the full roster picture, including schedule types, staffing models, standby time, and technicians moving on and off shift, so every assignment lands on someone available, qualified, and in zone. The Dispatch Agent re-optimizes against live signals: traffic, weather, client urgency, and case status update continuously, so a case is re-optimized when its urgency rises, replaced when a more urgent case arrives, or reassigned when another technician comes closer. The Orchestrator Agent coordinates dispatch and capacity so the two act as one decision layer.
Every autonomous decision is governed by six mechanisms: explainability, traceability, evaluation, autonomy levels, execution sandbox, and human-in-the-loop override.
The Results
People
Dispatchers moved from building plans manually to governing exceptions.
- Case assignment across every state runs autonomously; planners review exceptions and overrides instead of manually keeping the roster, the rules, and the map aligned.
- Each state's rulebook now lives in the system instead of planners' heads, so coverage no longer depends on who is on shift.
Resource
Every case now lands on a technician qualified to handle it, across all six work types.
- 15% less drive distance and time, with assignments computed on live distance and traffic.
- Standby time, shift changes, and zone gaps are planned for in advance, not scrambled around after they hit.
Cost
Penalty risk and operating cost fell together.
- 20% lower SLA penalty risk: fewer machines stay down long enough to trigger liquidated damages.
- 18% lower fuel spend from shorter routes across the full technician network.
Impactful Enterprise Stories: 360+ and Expanding
$1M+ in savings for a North American retailer's multimodal, ocean-to-store network
3X ROI for a global FMCG leader across 10 countries
$14M+ unused capacity uncovered for a Fortune 50 parcel leader in North America
75% growth in clinician visit capacity for a leading US home care provider
Know More About LOCUS Services?
Schedule a demo with one of our industry experts