The problem

What the customer was up against.

  • Field technicians arrived without enough information; 38% of jobs required a second truck roll because the wrong parts were carried.
  • Knowledge of complex grid components lived in the heads of 4-5 senior engineers approaching retirement.
  • Storm response involved 14 different call queues and a phone tree that often left customers without status for hours.
  • Asset failure prediction relied on simple time-based maintenance, missing high-risk units and over-servicing healthy ones.
The solution

What xyner built.

  • xyner agents on every field tablet pull the customer history, recent grid events, parts inventory, and the closest depot — then suggest the optimal repair plan before the truck rolls.
  • Specialist knowledge from retiring engineers captured in xyner's RAG knowledge base; agents apply that expertise on-call to any technician.
  • During storms, agents auto-prioritize tickets by safety risk and customer criticality, dispatching trucks via the most efficient route and updating customers proactively.
  • Connected to the SCADA system, asset-health sensors and weather feeds — agents now flag predictive failures 14 days ahead of actual breakdowns.
The outcomes

Measured impact.

31%
repeat truck rolls
reduction
42%
ticket-to-restore
faster
2,400
field technicians
live on platform
14 days
predictive failure
lead time
78%
fewer storm complaints
year over year
Executive summary

At a glance.

Situation

A regional integrated utility was running ~110,000 field-service calls per year with field-crew utilization at 58% and a customer-effort score in the lowest quartile. Outage communications were generic and late; field-crew arrival ETAs were unreliable; closing-out work-orders took post-shift admin hours that ate into crew productivity.

Intervention

Deployed a Field-Dispatch Agent and Customer-Comms Agent on xyner, with edge components at regional dispatch centres for sub-second response, and full integration to SCADA, OMS, the WFM system and the customer-information system.

Outcome

$10M+ annual savings (crew productivity, reduced overtime, avoided truck-rolls); field-crew utilization up to 73%; customer-effort score moved from bottom quartile to upper-middle; outage-comms accuracy and timeliness materially improved.

Industry

Energy & Utilities

A regional integrated utility (electricity and gas)

Scope

United States (multi-state)

Field-crew dispatch and customer communications during outages and field-service calls

Duration

8 weeks pilot, 5 months full rollout

From contract signature to full rollout.

Architecture

What the deployment actually looks like.

Field operations need low-latency dispatch decisions and resilience to backhaul issues. The deployment runs primarily in the utility's private cloud with edge components in regional dispatch centres for sub-second decisions and graceful degradation when central connectivity is impaired.

Field-Dispatch Agent

Selects the right field crew for each call based on skill, geography, current workload, parts availability and SLA tier; dispatches via WFM with route optimization.

OMS / SCADA integration

Subscribes to outage events; correlates customer reports against grid topology; identifies likely root cause and prioritizes restoration order.

Customer-Comms Agent

Drafts and dispatches customer outage and service-call comms — including ETAs, status updates, restoration confirmations — in customer-appropriate channels.

Edge components

Regional edge servers in dispatch centres provide sub-second dispatch decisions and continue operating during central-cloud connectivity issues; sync back when restored.

Work-order completion agent

After a field call, an agent drafts the work-order completion record from the crew's voice or photo input; the crew confirms in seconds rather than spending end-of-shift admin time.

Vulnerable-customer protections

Customer-comms respects the utility's vulnerable-customer policies; agents never use pressure language and always preserve clear paths to human support.

Implementation timeline

How the rollout sequenced.

An 8-week pilot covered one operating region before phased rollout across the utility's full service area.

Weeks 1-2

Foundations

Deploy private-cloud data plane; deploy edge components in one regional dispatch centre; integrate WFM, OMS, SCADA and the customer-information system.

Weeks 3-4

Agent configuration

Configure dispatch and comms agents against existing dispatch playbooks; complete first round of vulnerable-customer-policy testing.

Weeks 5-6

Shadow mode

Agents run in shadow alongside dispatchers; outputs reconciled; thresholds calibrated; integration to ITSM completed.

Weeks 7-8

Regional pilot live

Live in one region for field-service and routine outage response; metrics reviewed weekly with operations leadership.

Months 3-4

Full service-area rollout

Roll out across the full service area with per-region variations.

Month 5

Continuous improvement

Playbooks updated based on outcomes; autonomy thresholds tuned; storm-response playbooks introduced.

Governance & controls

How the deployment is governed.

Utility operations are tightly regulated and customer-safety-critical. The deployment was designed to strengthen the existing operational governance, especially around vulnerable customers and safety-critical communications.

OT-safety boundary

Agents never issue commands to grid-control systems; they only inform dispatch and customer-comms. OT actions remain human-controlled.

Vulnerable-customer protections

Customer-comms respects the utility's vulnerable-customer registry and policies; certain customers always receive enhanced support.

Safety-critical accuracy

Outage-communication content is gated for factual accuracy; never overpromise restoration; always include safety guidance for relevant incident types.

Regulatory reporting

Major-outage reports drafted automatically with the level of detail required by the state utility regulator.

Audit-grade trail

Every customer-comms message, every dispatch decision, and every work-order completion is captured for regulatory and internal audit.

What other enterprises can learn

Three transferable lessons.

Three lessons for other utilities considering agentic AI in field operations.

1

Edge-resilient design from day one

Field operations cannot stop when central-cloud connectivity drops. Edge components and sync-on-reconnect were design properties from week one, not retrofit.

2

Customer-comms is the single biggest experience lever

The dispatch optimization saved money; the customer-comms agent moved the experience needle. Both matter; their leverage is different.

3

Work-order admin is the hidden tax

Eliminating end-of-shift admin time gave crews back hours per week. The work-order completion agent paid for itself within the pilot.

Our crews stopped feeling like the system was working against them. The platform now gives them the right job, the right route, and the right paperwork at the right time.
VP Field Operations, US integrated utility

Reference call available through your xyner account team; the deployment was featured at a major US utility-industry conference in 2026.

Talk to a partner

Could the same outcome work in your environment?

Tell us your sector. A senior xyner partner will walk you through a tailored plan.