ePlanet · Fintech · multi-brand
ePlanet: one deployment running 15+ brand sites in 12+ languages
Another brand site is a configuration change, not a release. A new page is live in under ten minutes.
What we build
Most of this work never gets marketed, because it never faces the public. It is also the work that changes how a business runs on a Tuesday.
01 / 05
Operations dashboards
We find out how last week actually went on Monday, from three exports and someone’s memory.
Built to be read at a glance, drilled into when it raises a question.
One screen that pulls from every system holding a piece of the truth — the CRM, the finance tool, the site, the sheet nobody has migrated — and reconciles them into numbers your team stops arguing about. It gives them KPI cards, a hot-lead priority table ranked by intent, an engagement breakdown per lead, and a full activity timeline, all in the place they already work rather than a separate login they forget.
Cross-system reporting · KPI views · Live data · Role-scoped access
Internal tools that replace the department spreadsheet
One spreadsheet runs a whole department, only one person really understands it, and everyone is nervous when they take leave.
The spreadsheet survived because it was faster than asking IT.
It fails at the same points every time: two people editing at once, no validation, no record of who changed what, and no way to give someone access to their own rows without handing them everyone else's. We build the tool that does the same job properly — the forms your process actually needs, the rules enforced instead of remembered, an audit trail, and permissions granular enough that the tool can finally be given to the whole team.
Line-of-business tools · Validation & approvals · Audit trail · Permissions
Production tooling — files in, finished output out
A skilled, expensive person spends most of the day performing the same conversion by hand.
Doubling the volume no longer means doubling the team.
When the work is repetitive, high-volume and technical, it stops being a staffing question and becomes a tooling question. We build the pipeline that takes the files in as they arrive and returns the finished output, so your skilled people spend the day on judgement rather than on the same conversion by hand. The economics are the argument: cost per job stops tracking headcount and starts tracking compute, so doubling the volume no longer means doubling the team.
Batch pipelines · File processing · Volume economics · Custom rendering
Client and partner portals
Clients email us for updates that already exist on a screen inside the business.
Per-record permissions stop being a feature and become the product.
A login where each client, agent, supplier or franchisee sees their own data and only their own — status, documents, requests, history. It removes the daily round of chasing emails, and it does something a shared folder cannot: it makes the boundary between accounts a property of the software rather than a matter of care. This is where per-record permissions stop being a feature and start being the product, and it is exactly where site builders and their fixed connector lists run out of road.
Authentication · Per-account scoping · Document exchange · Self-service
Custom systems of record
We evaluated four products. Each covers about seventy percent of how we work — a different seventy percent each time.
A new brand launches as configuration, not a deploy.
Some businesses are genuinely shaped in a way no vendor has built for: several brands under one roof, multiple languages including right-to-left, teams that must be kept apart inside the same system, an approval chain that is the actual business. That is what we build: one deployment serving every brand site, multiple languages with full RTL, a page-builder section library, role-based access separating platform administrators from a single brand's editors, and per-domain SEO — so a new brand launches as configuration, not a deploy.
Multi-tenant · Role-based access · Multi-language & RTL · Workflow
Sometimes the right answer is to build nothing.
If you already pay for a platform that can do the job once it is configured properly, discovery will say so and there will be nothing here to build. And if nothing new needs a screen — if data simply has to move reliably between systems that already exist — that is operations work, not a software build.
Delivered work
ePlanet · Fintech · multi-brand
Another brand site is a configuration change, not a release. A new page is live in under ten minutes.
Mizanna Studio · Interior design
The drawings the studio already produces become the visuals a client needs in order to say yes.
A multi-product retailer · Retail
Social content, brand guidelines and the publishing queue are managed in Sanity, with a dataset per product.
Buy first wherever a product genuinely fits — it is cheaper and faster, and we will say so. Building is the right call when the process that makes you money is the part no vendor supports: several brands or entities in one system, teams that must be walled off from each other, an approval chain that is the actual business, or volumes where per-seat pricing stops making sense. The test we apply in discovery is simple — if the product forces you to change how you work in a way that costs you something real, that is when custom pays for itself.
That is the reason we write code rather than assemble a builder. Your CRM, your finance system, your CMS, your WhatsApp, your calendar, the internal database somebody wrote years ago — if it has an API, a database, a file drop or an export, we can connect it and keep it in agreement. Site builders such as Wix, Framer and Webflow are limited to the connectors their vendor has chosen to ship; when your requirement sits outside that list there is no way through it. Custom code has no such edge.
For a simple form-and-table tool used by a handful of people, often yes, and that is a fine place to start. No-code breaks in four predictable places: multi-tenancy, where several brands or entities need one system without seeing each other; permissions that have to hold at the level of an individual record; joining systems that disagree about the same customer; and volume, where per-seat or per-record pricing turns success into a penalty. Each of those is a structural limit rather than a settings problem, which is why the migration tends to happen later and cost more than building it properly would have.
No, and we would advise against it. The first phase is scoped to be the smallest thing that removes the pain you can name — usually one screen, one team, one process — running alongside what you have today. Once people are using it and it has earned trust, the next phase extends it. Each phase is quoted on its own, so the decision to continue stays yours.
A free first conversation, where we work out whether this is a build at all. If it is, a paid discovery goes into the process properly: how the work actually flows, what the systems already hold, who needs to see what. That produces a written recommendation and a phased quote. You own the discovery output either way — including the recommendation not to build, if that is the honest one.
Ownership is agreed before anything is written. On enterprise engagements you own everything from day one — source code, design system, CMS, data and infrastructure — deployed on your own infrastructure, with full documentation and handover if you want your own team to run it. On managed engagements we host, maintain and keep improving it, and your content, data and leads remain yours throughout.
Describe the process, who touches it, and where it breaks. We will tell you straight whether it needs building, configuring, or leaving alone — and if it needs building, what the first phase should be.
The first conversation is free. You leave it with a clear recommendation, whether or not anything follows.