MMEG.
Relay knowledge system

The citation-backed runbook corpus for micro-mobility repair.

Relay's knowledge system is the typed content substrate that shop operating systems, training programs, fleets, and insurers can license instead of building their own. Customer-facing runbooks compose from reusable atoms; every safety-critical step cites its source, so each repair leaves a verifiable record. The schema is the seam to the technician-facing app that runs these runbooks in the bay.

What it is licensed for

4 surfaces
Shop operating systems
License the runbook corpus instead of authoring your own. Customer-facing templates compose from reusable, versioned atoms.
Technician training
Every step maps to competency tags, so a curriculum can be assembled from the same content the bay runs.
Fleets and insurers
Each safety-critical step cites its source, so a completed repair leaves a verifiable record, not only an invoice.
The technician-facing app
The schema is the seam to the in-bay recorder: typed JSON and Markdown over a versioned API with content hashing.

Template preview

1 of 16 · structure only
ServiceEst.
TIRE / TUBE
Replace a flat or worn tire or inner tube.
template:tire-tube
~25 min
Intake · what the tech is asked
Which vehicle is this?
6 options
Which tire type? (sight-confirm where possible — defer if hub is opaque)
4 options
Composes 13 atoms
atom:intake-consult
atom:bay-readiness
atom:power-down-and-photo
atom:pre-service-check
atom:wheel-remove
atom:tire-tube-disassemble
atom:tire-tube-reassemble
atom:tire-swap-tubeless
atom:tire-swap-solid
atom:tire-swap-urethane
atom:wheel-install
atom:hub-cable-reinstall
atom:findings-handoff
LicensedThe compiled steps, specs, captures, and source citations are behind the licence. Sign in with an allowlisted account, or request a licence or demo to access the full corpus.
StageMMEG is pre-pilot. Estimated minutes are synthetic typical durations, not measured shop data (not yet validated). Wire schema version 0.11.0. The preview is rendered at build time from the corpus in data/.