MTN Guide
One platform, two parts. Guide, an agentic data engineer, reads your schemas, documentation, and API specifications, not your records. It builds MTN FieldMap, a map of what your data means across systems with the evidence behind each mapping, and keeps it current as those systems change.
Where MTN Guide lands
Beside your systems, not in the data path. Your stack keeps working as it does today.
The map, from schemas and docs
Re-checks only what a change touches
Routes to the right fields, with evidence
Live in days.
A new source is mapped against concepts the FieldMap already holds, so each system costs less than the last.
Maps fields using values, naming patterns, and standards.
High-confidence matches apply automatically.
Low-confidence cases route to a human, never guesses.
Change stays contained.
Conventional integration maintenance compounds with every new system. Mapping each source into its own layer keeps a change local to the connection it affects.
Self-healing pipelines adapt when sources change, with no manual rework.
Containment doesn't happen by accident.
Three ways a source can move, and what the FieldMap does about each.
New field detected and mapped
Renamed field matched by values
Type conversion applied automatically
Structure adapted with full audit log
Forward-only. Historical data stays stable.
Ready for what you deploy next.
A canonical layer your next model reads from on day one. Claude, GPT, Gemini, or your own.
Every model query lands inside the compliance perimeter
What it does, and what it doesn't.
MTN FieldMap is the layer between your sources and your stack. It doesn't try to be your stack.
It does
- Read schemas, docs, and API specs, no records required
- Map fields to shared concepts, with the evidence behind each
- Re-check only what a schema change actually touched
- Answer where data lives, for your team, your BI, your agents
It doesn't
- Replace your warehouse or BI tools
- Sit in your production data path
- Require changes to your source systems
- Hold your records, or return data as results
You find out when something moves.
A schema change re-opens only the mappings it touches, and conflicts surface the moment you ask.
A schema change re-opens only the mappings it touches, and conflicts surface the moment you ask.
See how this fits your stack
We'll walk through your sources, your warehouse, and your deployment plans, and show you where the map fits underneath.