Agentic service intelligence

The autonomous service layer that solves the first visit — before anyone rolls out.

AEDIL condenses the entirety of your service data — history, repair reports, technician notes, fault codes, parts, telematics, manuals, site data — into the finished decision: root cause, exact part, fix path. Telematics is one signal among many. The technician rolls out prepared. The first time.

Live in the pilot 88% diagnostic precision
The problem

The first visit doesn't fail at dispatch — it fails at diagnosis.

Dispatch decides who drives out. It doesn't decide whether the first visit fixes the machine. That's settled earlier — at diagnosis. And right there, before departure, the answer is missing today. Every single time.

1 in 3
first visits fails. A second van has to roll out — for a job that should have landed the first time.
first visits fail · industry-average FTFR
~€450
is what every return trip costs. Direct. Before the SLA clock even starts ticking.
per return trip · range €90–450
The entire service data memory was always there

The answer has long been sitting in the customer's systems — in the entirety of your service data: history, repair reports, technician notes, fault codes, parts and materials data, telematics, manuals and bulletins, customer and site data. Telematics is one signal among many. It just gets pulled together too late: in one person's head at a desk, long after the truck has rolled. You drive on a guess. The wasted second van is the price you pay.

The experience is retiring. The gut instinct that still saves the first visit today is walking out the door — and it isn't coming back.
The contract punishes you. SLAs turn every failed visit into a P&L hit — not just one more trip, but a penalty clause.
Service is where it's being decided right now — who leads, and who gets left behind.
Reactive
Diagnosis at the desk
The finding takes shape in someone's head — after the trip, once the technician is already standing in front of the machine. If the part is missing, a second van rolls out. That's how it runs today.
today · after the trip
AEDIL
Decided before departure
Root cause, exact part, and fix are ready before anyone sets off — distilled from the entirety of your service data already sitting in your stack. That's how it runs starting tomorrow.
before departure · dispatch-ready
The proof

Not a deck — hard numbers on real machines.

Live in the pilot · real field service, real visits
88%
Diagnostic precision
Top-3 hypothesis hits the installed part · across 650+ real cases
81%
Top-3 spare part nailed
The right part is in the van before you roll out
>90%
Proactive risk detection
Failures flagged before the ticket exists
Minutes
instead of manual research
What technicians piece together themselves · AEDIL does in minutes
Next in the pilot — first-time-fix rate as the primary KPI. We're now measuring what AEDIL moves on the ground.
The Job Pack · inside the AEDIL web app

A decision that reasons — and answers when you ask.

A machine signal fires. AEDIL pulls together every piece of your service data on its own — service history, repair reports, technician notes, fault codes, parts, telematics, manuals, location data — and condenses your entire service-data memory into a finished dispatch decision. Live in the AEDIL web app, queryable anytime. Telematics is one signal among many — not one sensor, the whole picture.

Intake
Asset ID#AX-2207
Symptomrecurring overtemperature under load
AEDIL pulls together every piece of service data on its own
Service history
Repair reports
Technician notes
Fault codes
Parts & materials
Telematics
Manuals & bulletins
Location & customer
Curio reasons
Logic Search Ranking condensing…
Delivered through your channels — AEDIL web app MS Teams PDF no new system
Next: autonomous parts ordering with SLA check

AEDIL does not pick the technician · 88% diagnostic precision on real machines

Real pilot case

The obvious sensor was wrong. AEDIL found the real root cause.

An anonymized case from Phase 1. The obvious sensor was already swapped — the steering faults stayed. AEDIL ranked the non-obvious causes · with confidence and historical evidence. The real fault sat right at the top.

AEDIL Web-App · Pilot case (anonymized)
Asset #…25 · Phase 1 · anonymized
Steering faults — stayed after sensor swap
Real case
Error codes
4371.2 4371.4 1710.067
Starting point
The obvious sensor was already swapped — the steering faults stayed. The obvious suspect was wrong. AEDIL prioritized the non-obvious causes — and nailed it.
Causes · ranked with confidence
5/5
Parameter misconfiguration (P415) — Parking-Brake parameter
4/5
Internal ribbon cable — JetPilot Internal Harness
3/5
Bearing — Four-Point Bearing Gear Grub Screw
Backed by
2 precedent cases with an identical signature — verbatim from the history #…96 · #…78
The first visit landed. Because the diagnosis put the real fault up front — not the obvious one.

Real case, Phase 1 · customer and asset anonymized · error codes and cause ranking verbatim from the pilot

Runs on your stack

A layer over your stack — no rip-and-replace.

The Enterprise Adapter reads the entirety of your service data across every system you already run — ERP, FSM, CRM, telematics, knowledge base — not just one source, and lays the diagnosis layer on top. No migration. No second database. Live in weeks, not quarters.

System architecture
Three layers, one stack

AEDIL extends your systems — it replaces nothing. The adapter sits between the agents and your ecosystem, reads the entire service-data landscape and translates both ways.

Layer 1Agents
Porta Curio Aquila
Layer 2Enterprise Adapter
Enterprise Adapter
Layer 3Your ecosystem
SAP Salesforce IFS Jira ServiceMax
Today the adapter reads live · autonomous write-back into SAP/FSM comes next.
The platform

One engine, multiple entry points — the lead grows with every job.

One engine, from machine signal to diagnostic decision. It runs live in three places today and gets more autonomous with every deployment. Multiple entry points, one data lead.

Machine signal Diagnostic decision growing autonomy
Live today
On-Site · Maintenance
Live
Briefing · 06:00
Top 5 at-risk machines
>90 % recall
1
Asset A-204
Temperature sensor drifting — calibrate before shift.
94
2
Asset F-118
Bearing vibration rising — endoscopy recommended.
88
3
Asset C-077
Coolant flow near limit — check the pump.
71
4
Asset B-339
Hydraulic pressure fluctuating — keep watching.
52
Before the ticket · >90 % proactive risk detection
Field Service · Pre-Dispatch
Live
A
AEDIL
Teams · to technician
05:58
Job Pack · sample case
Drive unit — overtemperature under load
CauseSensor drift on the temperature sensor
Part#TS-4471 · temperature sensor
Confidence 87 %
Open Job Pack
Before the visit · 88 % diagnostic precision
Hotline · "Ask AEDIL"
Live
Mast judders on lowering — which part do I check first?
AEDIL
Case #AX-1903 · anonymized
Ask about this case
On the call · 81 % top-3 spare part

The same engine, more autonomous stage by stage. This is the roadmap — here's where the engine lands next.

Self-service recommendation to the end customer Escalation path to the dealer Autonomous parts ordering · SLA check SAP / FSM integration The self-closing learning loop
One system · three nodes
Three agents, one line — every job closes the learning loop.
Porta
Signal intake
Catches the machine signal — before the ticket.
Curio
Diagnostic engine
Ranks the cause and the exact part for the decision.
Aquila
Feedback loop
Feeds the visit outcome back — sharpens the next diagnosis.
Closed loop
Aquila → Curio · every visit makes the next decision better
The data lead grows with every job — that's the moat.
Diagnosis is the beachhead, not the goal. Every visit closes the loop through Aquila and sharpens the next decision. A later entrant never catches up on this lead.
The vision

Diagnosis is the bridgehead.

Today AEDIL nails the first visit — before anyone rolls out. That's only the start. On the diagnosis layer grows the autonomous service layer: it doesn't recommend. It acts.

The next step
We're building the autonomous service layer for industry.

Two minutes. From one machine signal to a dispatch-ready decision — root cause, exact part, the fix.

Live in the pilot · no new system · runs on your stack
Live today
  • 88%Diagnostic precision — the right root cause, on the first pass.
  • 81%Top-3 spare part — the right part among the first three.
  • >90%Risk detection — the failure flags itself before a ticket even exists.
The manual research AEDIL does in minutes

Service is where the lead is decided now. We win the decision where it's made — before the visit.