Solution
A shared workbook is usually where good business software begins. When it starts locking, recalculating for an hour, or hiding data that some people should not see, we turn its proven logic into a real-time, role-scoped web application. Senior in-house engineers. Conventional React and TypeScript. You own the code.
Who This Is For
Sales or revenue models in a workbook that several people edit, where one change means waiting for everything to recalculate. Alcon: 30 to 60 minutes per change before the app.
Sensitive data that different roles need to see in different ways, with no clean way to hide a tab from the wrong person. Rees-Jones Holdings: seven roles, one dataset.
Numbers exported from a legacy system, pasted into sheets, and emailed around. EZ Storage: 15 facilities, real-time reporting on a custom dashboard.
Contacts, contracts, and sales targets spread across workbooks and inboxes. MJEC: 125+ manufacturer lines, now one governed system.
Not for
A workbook one person uses once a quarter, or a sheet that already works and simply needs a chart. We will tell you when the spreadsheet should stay a spreadsheet.
What We Build
The rows, tabs, and lookups become modeled objects with history. Products, forecast versions, employees, territories, and sales figures live in PostgreSQL, MSSQL, or Firestore, not in cells.
Who can see and edit what is designed into the application, not into a protected sheet. Each role gets its own interface and its own slice of the data.
Projections and totals update as the data changes. The hour-long wait for a shared file to finish disappears, and so does the version confusion that came with it.
Dense model output becomes interactive visuals that support decisions, and the exports people still need (PDF, CSV, a branded document) come straight from live data.
Every edit is attributed and recoverable. A bad paste no longer destroys the only copy.
How We Work
We start from the workbook, not from a blank specification. The logic it encodes is usually right. The container is what has to change.
Every tab, lookup, and macro, plus the people who touch them and the parts of the process that hurt. We map what should stay and what should change.
Products, forecast versions, employees, territories. Then who sees which of them. Role design is architecture, decided before any screen is drawn.
Stakeholders click their own data on a shared staging environment, not a slide. The workbook keeps running alongside until the app is trusted.
The spreadsheet stops being the system of record. Most clients keep a quarterly release cadence with us after launch, as MJEC does.
Work
Shared Excel workbooks that took 30 to 60 minutes per change, replaced with a real-time forecasting application built alongside the Alcon data science team.
View case study5-year
forecast horizon
Manufacturer and partner data spread across spreadsheets and disconnected tools, consolidated into one governed internal app plus a field brochure tool.
View case study125+
manufacturer lines
A custom internal dashboard giving a 15-facility operator real-time reporting and automated SEO audit reports on their own numbers.
15facilities
View case studyA firm-wide intranet in daily use by every employee, with practice-gated content and role-based access, maintained since 2014.
10+years in daily use
View case studyLive demo
Coming soonAn interactive demo on synthetic data: pick a role, edit an estimate, add a note, and watch a five-year projection recompute in real time. Modeled on the Alcon engagement without any Alcon data.
Synthetic data. No sign-in.
Technology
Application
Data and hosting
FAQ
No. The formulas, lookups, and rules are usually the most valuable thing in it. We translate them into application code with tests, then verify the app against the workbook before cutover.
Yes. Exports, CSV downloads, and generated PDFs come straight from live data. What changes is that Excel stops being the place where the truth lives.
Role-based access is designed into the data model and the interface. Rees-Jones Holdings needed seven roles on one compensation dataset, each with a different view and different permissions.
It keeps running. We ship working software in two-week slices to a shared staging environment, and the team moves over when the app is trusted.
No. You own a conventional React and TypeScript codebase that any senior web developer can maintain. The difference is ownership and fit.
When one person uses it, the data is not sensitive, and nobody is waiting on it. We will say so in the first conversation.
Keep Going
The product shape: dashboards, RBAC, admin consoles.
SSO, systems of record, agent-ready handoff.
Start A Conversation
Send the workbook or a screenshot, tell us who uses it and what goes wrong. We will tell you what should stay, what should change, and whether a custom tool is worth it at all.
Start a conversation