AttomikATTOMIK
WORK / MATPRO
CLIENTACTIVE · BUILDING

From spreadsheets and WhatsApp to a single source of truth.

A Mexico City waste management company had no system of record. Attomik built one, on the principle that a number that's never stored can never drift from its source.

RELATIONSHIPCLIENT
SCOPEOperations platform: catalogs, fleet, billing, receivables, reporting
CATEGORYWaste management · Mexico City
LARGEST ACCOUNTGrupo Posadas — 9 hotel properties, 9 billing entities
A Matpro waste collection truck on route in Mexico CityMATPRO · OPERATIONS[ SOURCE ]

THE CLIENT

Materiales y Procedimientos de México is a Mexico City waste management company serving hotel groups, retail chains, and industrial clients.

Their largest account, Grupo Posadas, spans nine hotel properties billed through nine separate legal entities.

THE PROBLEM

Matpro had no system of record. Operations lived in Excel, WhatsApp, and PowerPoint. Every month the administrative team rebuilt a client-facing deck by hand: pulling volumes from fuel sheets, retyping totals, reconciling receivables against invoices spread across multiple billing entities.

The errors were predictable and always in the same place: the hand-typed summary slides. Numbers drifted from their sources, and nobody could tell which version was right.

WHAT WE BUILT

A unified internal operations platform, designed around the person who uses it every day (the administrative manager), and organized by what she does rather than by data type: capture, review, report, configure.

01CatalogsClients, billing entities, sites, and containers, with commercial clients modeled separately from the legal entities they invoice through.
02FlotaFleet and vehicle document tracking with expiry visibility.
03ServiciosFive pricing models, including per-volume billing that surfaced only once we analyzed real Posadas invoices.
04CobranzaReceivables across all billing entities.
05VolúmenesMaterials and volume tracking per site.
06CierresFormal monthly period close.
07Panel y ReportesLive dashboard and year-to-date views that replace the manual deck.

THE ARCHITECTURAL DECISION THAT MATTERS

No stored aggregates, anywhere.

Every total, every dashboard figure, every report number is computed at read time from the underlying detail rows.

This was not a technical preference; it was a direct response to the root cause of the client's data problems. If a rollup is never typed and never stored, it can never drift.

Materials sorted and staged at a waste processing facilityMATPRO · FACILITY

HOW WE GOT THERE

We loaded Matpro's real operational history before designing the final screens: clients, sites, vehicles, invoices, payments, fuel loads, container movements, service events. That analysis pass exposed schema gaps that no requirements document would have caught, including the per-volume pricing model and the multi-entity billing structure.

We then reconciled our computed volumes against what had actually been invoiced and matched on the large majority of Posadas sites, which validated both the data model and the extraction.

Before and after

BEFORE

System of recordNone — Excel, WhatsApp, PowerPoint
Monthly client deckRebuilt by hand every month
RollupsHand-typed, could drift from source

THE ATTOMIK LAYER

NO STORED AGGREGATES, ANYWHERE

AFTER

System of recordOne platform — ops, billing, fleet
Monthly client deckProduced without manual rollups
RollupsComputed at read time, never stored
TYPED BY HANDCOMPUTED LIVE

WHERE A ROLLUP LIVES

RESULTS

One place where operations, billing, and fleet data live together.

Monthly reporting that the team produces without manual rollups.

Totals that are always traceable back to source rows.

A platform built on Attomik's starter foundation, so the client got production-grade auth, permissions, audit trails, and theming from day one, and effort went into their domain instead of scaffolding.

WHAT COMES NEXT

Client-facing sustainability and waste reports: automated monthly and annual decks per Posadas property, generated as PDF and HTML email directly from platform data.

Replacing seed data with the team's live Excel workbooks (fuel, receivables, payables, containers) and hardening the screens against real usage.

Role refinement as the team grows into the system, including who can close a month.

Longer term: operator field capture and a client portal, so data enters once at the point of service and clients see their own reporting without a deck being sent.

WHY THIS IS REPEATABLE

The Matpro pattern applies to any small operations team running on spreadsheets and chat: analyze the real data first, model the domain honestly (including the messy parts like multi-entity billing), derive everything from source rows, and design around the one person who lives in the tool.

START A CONVERSATION

Which operation is still running on spreadsheets and chat?

Start a projectSee enterprise AI