SIGNATURE · When the spreadsheet becomes a risk

Platform 90

Your business application or SaaS product in production in 90 days. Frozen scope, firm date, and code you fully own.

Format
Fixed price
Duration
90 days
Tier
SIGNATURE
For
SMEs and founders

IN 1 MINUTE

  • You get: an application in production, with real users on it, on day 90.
  • Investment: fixed price, scope frozen at the end of week 2.
  • Guarantee: the date is held, or the additional weeks are on me.
  • Ownership: code, infrastructure and documentation fully transferred. No dependency on NOKARIS.

Our commitment

If the production date is missed, the additional weeks are on me.

I can hold that guarantee because I freeze the scope rather than promise to move faster. Requests arriving mid-project are neither refused nor silently absorbed: they are quoted and dated as v2. That is exactly what separates a project that lands from one that drifts.


Who it is for

Two situations, one answer.

  • A critical process held together by spreadsheets. Quotes, case files, scheduling, production tracking: it still works, but only one person knows how, and a bad copy-paste costs a day.
  • A SaaS product to bring to market. You have the domain knowledge and the first users in mind. What you lack is someone able to design, build and operate the platform — not just write code.
  • An available decision-maker. One hour a week to arbitrate. Without it, no scope stays frozen and no date holds.
  • This is not for you if the need cannot yet be articulated, or if the scope must stay open. In that case start with a discovery engagement — or take AI Quickstart to address the most painful process first.

Method

Ship in 90, applied to custom software.

  1. Weeks 1-2 · Discovery

    Workshops with you and your field leads. Domain model, critical journeys, key screens, architecture decisions, migration of existing data. Scope is settled and signed at the end of week 2 — the only moment it can still move.

  2. Weeks 3-10 · Build

    Increments shipped weekly to an accessible staging environment, with a Friday demo. Hexagonal architecture: business logic stays independent of the database, the payment provider and the host. Automated tests and continuous integration from week one, not at the end.

  3. Weeks 11-12 · Production launch

    Infrastructure as code, monitoring and alerting, backups, real data migration, load testing on critical journeys. Training for users and for whoever will operate the tool day to day.

  4. Day 90 · Handover

    Repository, access, infrastructure and operations documentation handed over. A costed v2 roadmap built from what was deferred during discovery and what real usage has since revealed.


Thirty minutes is enough to know whether your need fits a 90-day window — and to tell you plainly if it does not.

Book 30 min →

What you keep

Everything. Including the ability to do without me.

Code

Full Git repository, history included. No proprietary NOKARIS component.

Infrastructure

Described in Terraform on your own cloud account. Reproducible elsewhere.

Tests

Automated suite and continuous integration. The next developer knows when they break something.

Documentation

Architecture, operations, recovery procedures. Written for someone arriving two years from now.

This clause is not generosity, it is a design constraint. A deliverable you cannot take over is a deliverable whose quality is never tested.


Why now

A critical spreadsheet never breaks at a convenient time.

The cost of a process held together by shared files appears nowhere in the accounts. It shows up the day the person who knew how it worked leaves, when two versions diverge unnoticed, or when a client receives wrong data. By then the rebuild happens under pressure, and therefore badly.

On the product side the window is different but just as real: a validated idea that does not reach production within the year gets overtaken, or loses the context that made it relevant. Ninety days is the span that still lets you learn from the market rather than guess it.


The proof

Platforms already running in production.

  • Kestrelia — a flight-fare monitoring SaaS designed, built and operated solo, live in the North American market. Asynchronous workers, three providers behind a single interface, AWS infrastructure described in Terraform. The code is open for inspection during the discovery call.
  • JobPilot — a job-search automation platform: Next.js 16, hexagonal NestJS API, Supabase, Turborepo monorepo. Built with the same method, with an end-to-end test suite and continuous integration.
  • goclock.ca and saas.mtaxapp.ca — platforms in operation, built and maintained under mandate.

That is the difference between a supplier who writes code and someone who has already carried responsibility for a product in production, on-call included.


Quebec subsidies

Custom development is the most favourable expense base.

Digital development fees incurred in Quebec are exactly what the PSCE and CDAE programs target. On a Platform 90 engagement, this is the line item most likely to be covered. Our Tremplin Subventions offer builds the application file and is credited if you continue with NOKARIS.

See Tremplin Subventions →

Common questions

Five honest answers.

Ninety days — is that not too short for a product?

It is short for a complete product, and that is deliberate. Ninety days is enough to put a v1 in production that does one thing well, with real users on it. Whatever does not fit goes to v2, decided during discovery and written down. The usual alternative — an open scope over twelve months — mostly produces projects that never reach production.

Who owns the code?

You do, entirely, from the final payment. Git repository, Terraform-described infrastructure, environment variables, operations documentation: everything is transferred. No proprietary NOKARIS component, no hidden subscription, no dependency on me to evolve the product.

How do you hold a firm date on custom work?

By freezing the scope at the end of discovery, not by working faster. The first two weeks exist precisely for that. What is signed at the end of week 2 is what ships on day 90. Any later request is quoted separately and goes to v2 — never silently absorbed, never billed as an overrun.

What if our data already lives in spreadsheets or an old tool?

That is the most common starting point, and migration is in scope: modelling, cleanup, a replayable migration script, a consistency check before cutover. What is not included is rebuilding history that does not exist or cannot be used — and we say so during discovery, not on day 80.

How is this different from NOKARIS Embedded?

Embedded sells capacity: one or two senior developers on your roadmap, by the month. Platform 90 sells an outcome: an application in production on an agreed date, at a fixed price. If you know what you want built, take Platform 90. If you have a continuous roadmap and a team to reinforce, take Embedded.


Ready to start?

A 30-minute discovery call. We check whether your need fits 90 days, and I tell you plainly if it does not.