SaaS

We build SaaS products, and we run our own

Multi-tenant architecture, usage metering, billing, onboarding and the admin tooling underneath it all. The same problems we hit in our own products, built into yours before a pricing change or an enterprise contract forces the rebuild.

The sector

The half customers never see

A SaaS product is two systems wearing one name. There is the application customers log into, and there is the machinery that makes it a business: tenant separation, plans, metering, invoices, trials, invitations, seat changes and the admin screens your own team lives in. Customers rarely notice the second one. It is what decides whether the first survives growth.

That machinery is what gets deferred. Billing is a single price and a checkout page until a customer asks for annual invoicing. Metering is a nightly count until finance cannot reconcile it. Onboarding is a signup form until trials stop converting and nobody can say at which step. Each shortcut is reasonable on the day it is taken and expensive on the day it is unwound.

Villaex runs its own SaaS products, so we have paid for these decisions with our own time. NexaVoxa is multi-tenant, meters platform minutes and model usage, and bills against that usage. The counters behind those invoices are reconciled, not assumed. We build client platforms with the same care, because we operate what we build.

What makes it hard

The constraints that shape saas systems

Tenancy is decided before tenants exist

The isolation model gets chosen early and constrains everything after it. Shared tables, schema per tenant or database per tenant each carry a different migration, backup and support story that you will not get to revisit cheaply.

Usage has to be countable

Billing on usage means every billable event must be captured, deduplicated and replayable. A dropped event is revenue you never invoice. A double-counted one is a refund, an apology and a customer who now checks every line.

Trials decided in the first session

Running our own trials, what we see is that anything between signup and the first useful result costs you the account: an import that fails quietly, a settings page with no defaults, an invite that needs an admin.

Churn signals never meet

The seats removed at renewal sit in the billing record. The admin who stopped logging in sits in product events. The third reopened ticket sits in the support inbox. Nothing puts the three together.

What we build

Systems we ship for saas teams

Multi-tenant application platforms

The tenant model, isolation enforced where the queries run, per-tenant configuration and feature flags, roles and invitations, and an internal console where your support team can see one customer's workspace without guessing.

Usage metering and billing

An event pipeline that turns raw usage into billable counters, subscriptions and prepaid balances in Stripe or the processor you already use, proration on plan changes, invoices, and a pricing page a customer can act on without emailing sales.

Onboarding and activation flows

Signup, workspace creation, data import, teammate invites and guided setup, with the events instrumented at each step so you can see exactly where trials stall instead of arguing about it.

Account health and churn reporting

A per-account health record assembled nightly from product events, billing and your support inbox, with the drop rules written down so the same account does not look healthy in one screen and at risk in another.

AI features inside your product

Retrieval over a customer's own documents, assistants that draft the next step inside a screen people already use, and agents that call your existing endpoints, built so one tenant's data never reaches another tenant's answers.

Proof

Screens from systems in production

app.nexavoxa.com/agents/dental-appointment-assistant
NexaVoxa agents page with an agent configuration form
Voice AI Platform · Multi-tenant SaaS

Build a phone-answering AI agent in an afternoon, not a quarter

NexaVoxa is our voice AI platform. Teams create an agent, choose how it runs (a single prompt or a guided flow), decide who speaks first, write the welcome message and instructions, connect a phone number and knowledge base, and talk to the agent from the browser before it ever takes a real call. Multi-tenant, usage-billed, with analytics, call logs and collected data built in.

app.nexavoxa.com/billing
NexaVoxa billing and usage with pricing plans
Usage Billing · SaaS Monetisation

Pay as you go billing with a credit balance and an enterprise tier

NexaVoxa bills per minute of platform time plus model usage, tracked against a prepaid credit balance, with an enterprise plan for managed deployments. The billing screen shows the current plan, what is included, and how to move up, so pricing is never a support ticket.

FAQ

Questions we get before we start

Still unresolved? A 30-minute conversation with an engineer usually settles it faster than another page of copy.

Yes. That is the usual starting point. We read the codebase, run the deployment ourselves and write down what actually breaks before proposing anything. From there we work in small releases against your existing pipeline, so the product keeps shipping while the tenancy, billing or onboarding work lands behind feature flags.

By what your customers will ask for. Shared tables with a tenant column are simplest to operate and migrate, and they are the right default unless something forces otherwise. A schema or database per tenant is worth the operational cost when contracts demand hard separation, per-customer restore, or data kept in a particular region. We write the decision down with its trade-offs so it can be revisited.

You own the code. It sits in your repositories and your cloud accounts from the first commit. We work next to the engineers you already have, in the same pipeline and the same review process, or we staff a team where you have none. Handover is written as we go: runbooks, architecture notes and a walkthrough. Who carries the pager after launch is agreed in the contract, not assumed.

We build the evidence a review asks for. Audit logs that record who changed what and when, access review exports an admin can produce without an engineer, encryption at rest, and change management enforced in the deployment pipeline rather than in a document. We do not sell the audit itself. Villaex holds PCI DSS compliance and ISO 18295 certification, and we say nothing beyond that.

Building something for saas?

A 30 minute conversation with an engineer is usually enough to scope it honestly.