The answer is already in your telemetry.
Relay makes sense of it.

Relay reads the telemetry you already collect — the gateway, the modem, the node — and turns it into a plain-language cause: the subscriber can fix it, the agent can verify it, the tech knows what to bring.

Existing telemetry
ROUTER · GATEWAY
MODEM · ONT
CMTS · NODE
Diagnosis
Verdict · this one's on us
· conf 0.00 · owner: plant
Who gets it
SUBSCRIBER
“The line coming into your house is weak. That's ours to fix.” A crew is assigned for Thursday morning.
SUPPORT AGENT
Open · 00:41Relay
Most likely causeDrop attenuation
rx power-31.4 dBm
resets6 / 48h
confidence0.91
FIELD TECHNICIAN
Ranked cause
Drop attenuation91%
Tap or splitter6%
In-home coax3%
Probable parts
  • Drop cable 150ft
  • F-connectors ×4
  • Ground block

The problem

One call becomes four, and nobody's any closer to a cause.

You already collect the telemetry that would explain what's wrong. It just never reaches the people who need it — the subscriber, the agent, the technician — in a form any of them can use.

  1. 01

    The subscriber

    describes a feeling — slow, dropping, broken — not a cause. There's no way to say what's actually wrong.

  2. 02

    The agent

    works from a script, not a diagnosis. Twelve questions in, they're still guessing.

  3. 03

    The technician

    rolls a truck without knowing what to bring. Sometimes there's nothing to fix at all.

  4. 04

    The subscriber, again

    calls back next week. Same issue, same guesswork, one more ticket.

How it works

One diagnosis engine, rendered three ways.

  1. 01

    Existing telemetry

    Relay reads what your gateway, modem, and node already report — rx power, reset counts, node cohort. Nothing new to install.

  2. 02

    Diagnosis

    A rules-first classifier ranks the likely cause against five known failure patterns and states a confidence, not a guess.

  3. 03

    Who gets it

    One verdict, rendered three ways: a plain-language answer for the subscriber, a diagnosis for the agent, a parts list for the technician.

For co-ops, municipals, and independents

Built for operators whose subscribers are also member-owners.

Fewer trucks that find nothing wrong

Dispatch with a ranked cause and a parts list, not a guess. The technician knows what to bring before they leave the lot.

One incident, not twelve tickets

The agent sees the diagnosis before asking a single question — verify it, don't rebuild it from scratch on every call.

Honest, even when it's on you

Relay names the provider's own fault when that's the true cause, in language the subscriber actually trusts.

Yours, not ours

White-label in your own app, portal, and support workflow. Subscribers never see the Relay name.

Why this matters

The cost of a guess, in industry numbers.

≈$1,000

fully-loaded cost of a single truck roll once labor, fuel, and overhead are counted.

Technology & Services Industry Association, via Blitzz

≈25%

of truck rolls are non-value-added — the technician arrives to find nothing to fix.

isEMAG, "The Bane of the Unnecessary Truck Roll"

12–20%

typical repeat-contact rate for ISPs in normal operations — the same issue, more than once.

Umbrex, Repeat Contact Rate guide

Relay is pre-launch. We don't have Relay customers to quote yet — we're building the first version in conversation with operators now. That's what the pilot list below is for.

Pilot list

Talk to us about a pilot.

Relay is pre-launch. We're building the first version with a small group of co-ops, municipals, and independent providers. Leave your email and we'll reach out.