Home / Client Stories / Enterprise Operations Platform
Flagship Client Story · Workflow Modernization

The operations platform its acquirer couldn't replace

We spent eight years building a custom enterprise operations platform for a U.S. medical equipment service organization. When the company was acquired, the plan was to migrate everything onto a standard enterprise system. The plan changed — because the platform modeled the business better than anything off the shelf could.

700k+ work orders processed
300k+ instrument repair orders
2,000+ active users
150+ field specialists, offline
8+ years in production
The starting point

A business that had outgrown its systems

The client services medical equipment for hospitals, surgical centers, and healthcare systems across the U.S. — depot repair, field repair from mobile vans, loaner devices, vendor coordination, quality control. Their ERP handled the money. Nothing handled the work.

Before
Repair tracking spread across spreadsheets and email
Paper contracts interpreted manually before every quote
Inspection findings buried in notes and inboxes
Customers calling for every status update
No single system describing the full repair lifecycle
After
One platform owning every operational workflow
Contracts as live business rules inside quoting
Template-driven QC with photos and measurements
Customer & vendor self-service portals
Complete audit trail on every device, every repair

A standard CRM could hold the accounts. The ERP could hold the ledger. Neither could model a medical device moving through intake, inspection, quote, approval, repair, final QC, and shipment — with a loaner keeping the hospital running in the meantime. So we built the system that could.

What we built

One platform, every workflow, four audiences

The platform grew from CRM and dashboards into a custom vertical ERP — the operational system of engagement, integrated with the client's existing ERP as the financial system of record.

Existing ERP Financial system of record APIs · scheduled sync Enterprise Operations Platform Operational system of engagement · 500+ tables · 12 modules Work Orders Quality Control Operational CRM Contracts Loaner Assets Reverse Logistics Self-Service BI Preventive Maint. Fleet Analytics Data Quality Ops Business Reviews Customer Portal Hospitals & health systems Vendor Portal Manufacturers & repair partners Offline Field App 150+ techs in mobile vans Shipping Labels & tracking
System context — one shared workflow engine, exposed through role-specific experiences. Diagram recreated for publication; no client systems shown.

One business object at the center

Every repair — intake, inspection, quote, approval, repair, final QC, shipment — lives on a single Work Order. Every department works on the same record from its own perspective. No duplicate processes, no swivel-chair data entry.

Portals, not separate systems

Customers create and track repairs; vendors update outsourced work. Both portals expose controlled views of the same internal workflow — one engine, one audit trail, zero synchronization headaches.

The ERP stayed where it belonged

We never tried to replace the client's ERP. It remained the financial system of record while the platform owned operations — connected through APIs and scheduled synchronization. Lower risk, faster delivery, preserved investment.

Eight years of evolution

Built module by module, never a big-bang rewrite

  1. 2017

    Foundation

    CRM, dashboards, reporting, and ERP integration groundwork. The goal was operational support — not yet an ERP replacement.

  2. 2018

    The work order becomes the center

    Work order management, structured quality control, the customer portal, and the offline field service application. The platform became the daily workspace for operations.

  3. 2019

    Enterprise growth

    Vendor portal, loaner asset management, contract management, preventive maintenance, fleet analytics. Business processes became fully interconnected.

  4. 2020–22

    Scaling

    Working-vs-history data architecture, background archival, query optimization, deeper ERP synchronization. The database passed 500 tables while screens stayed fast.

  5. 2022–25

    Mature platform

    700,000+ work orders, 2,000+ users, millions of history records — plus a full migration of the platform to a modern, long-term-supported .NET stack.

  6. Today

    Still in production

    The platform continues to run the business years after active development ended — the quietest, strongest endorsement enterprise software can get.

Under the hood

Three decisions that made it last

Longevity wasn't luck. It came from architecture decisions that respected how the business actually operates.

Decision 01 · Offline-first field service

Designed for hospital parking garages, not perfect Wi-Fi

Field technicians repair surgical instruments from mobile vans — often in underground garages with no signal. A web app was never going to work. We built an offline-first desktop application on a local database: technicians sync in the morning, work all day disconnected, sync when they're back online.

The key innovation: synchronization transfers the entire local database as one atomic snapshot instead of dozens of fragile API calls. Either the whole sync succeeds, or nothing changes. Interrupted connection? Just retry. No partial states, no lost work — ever.

150+
field specialists
50–70k
syncs per year
0
partial-sync data states

Read the full solution story →

Field laptop Local database works fully offline Central platform Processes snapshot builds fresh copy 1 · upload compressed snapshot 2 · download fresh database All-or-nothing · atomic
Atomic snapshot synchronization — the whole database travels as one unit.
Decision 02 · Lifecycle-aware data architecture

Fast at 700,000 work orders — by design

Most systems slow down as data accumulates. We recognized early that operational and historical data have completely different jobs: users work on today's 7,000–10,000 open orders, while hundreds of thousands of closed orders exist for history, audits, and reporting.

So the database separates working tables (small, hot, heavily indexed) from history tables (read-mostly, unlimited growth). Background jobs continuously archive closed work. The result: screens stayed sub-second for eight years while the dataset grew into the millions of records.

700k+
total work orders
7–10k
active at any time
500+
database tables

Read the full solution story →

Working tables open orders only fast & small History tables closed orders · millions of records unlimited growth archive background jobs · no user impact
Closed work moves to history automatically — today's screens only carry today's load.
Decision 03 · Reporting as a platform

Business users answering their own questions

Instead of developers hand-building every report, we built a metadata-driven reporting engine. Developers expose curated datasets; business users build, filter, pivot, save, and export their own reports — no release cycle required.

The showpiece: automatic PowerPoint generation. Account managers select a customer and a period, and the platform assembles a presentation-ready business review from live operational data. Hours of manual preparation became minutes — and every meeting ran on current numbers.

1
reporting engine, all modules
Self-serve
reports, dashboards, pivots
PPT
generated from live data

Read the full solution story →

Curated datasets Report engine filter · pivot · chart save · export built by business users Excel Dashboards PowerPoint
Developers expose trusted data once; the business reuses it forever.
The outcome

The proof no marketing metric can match

When the client was acquired, the plan was to retire the platform and migrate onto the acquirer's standard enterprise system. After evaluation, the custom platform stayed — and half of the combined organization kept running on it.
It wasn't preserved because it was fashionable. It was preserved because it modeled the business correctly.
8+
years in continuous production
700k+
work orders processed
2,000+
users across every department
1
planned replacement — survived

Figures are production-scale numbers from the engagement. Client identity withheld under confidentiality; diagrams recreated for publication.

What this means for you

The same work, productized

Mapping the repair lifecycle before writing code — understanding how the business actually ran, department by department.
Operational Intelligence Assessment →
Eight years of architecture decisions — ERP boundaries, portals, offline strategy, scale planning — made alongside the client's leadership.
Fractional CTO & Technology Leadership →
Replacing spreadsheets, email, and paper with integrated workflows — work orders, QC, contracts, logistics, portals.
Workflow Modernization →
Clean data foundations and a self-service BI layer — the prerequisite for the AI capabilities we build today.
AI Enablement →

Your operations deserve software that outlives the slide deck.

Start with an Operational Intelligence Assessment — the same understand-the-business-first approach that made this platform irreplaceable.

Book an Operational Intelligence Call