Product rebuild · Switzerland
Web application and SaaS rebuild.
I rebuild web applications, SaaS products and business software that have aged — or prototypes that found customers without the architecture to keep them. The rebuild starts from the domain model: data re-laid and migrated, rules rewritten as tested code, payments, roles and administration rebuilt, then a cut-over without downtime. Fixed fee, on a written scope, usually CHF 25’000–80’000.
Updated
In brief
- Where
- Zürich, Switzerland — remote across Switzerland and Europe
- For
- Web application, SaaS, business software, prototype turned product
- Range
- Usually CHF 25’000–80’000, fixed fee; minimum CHF 20’000
- Typical timeline
- 5–10 weeks; beyond CHF 80’000, successive fixed-fee phases
- Entry point
- Paid diagnostic CHF 2’400 when the scope is not set, credited against the rebuild
- Built and run
- Atamatax · Horvera · Vellaci (in production)
A rebuild is not a new theme. It is the moment the accumulated debt gets paid: a data schema that no longer matches the business, rules scattered across screens, manual fixes in the database, a team that no longer dares touch the code. The rebuild puts the business back at the centre; the interface follows.
No rebuild done for a client is published yet. What I can show is my own products, built around their domain model — versioned rules, results that stay reproducible after a rule changes, data in one place: the architecture a rebuild has to produce.
01 — Signals
When a rebuild becomes necessary
The symptoms that come back in almost every product that has aged. One is not enough; three together are.
- Every change breaks something else
- Logic is copied across screens, changes hit unexpected places, and tests are missing — or do not test what matters.
- The business rules live in two heads
- Pricing, eligibility, statuses: when the person who knows is away, support stops.
- Data gets fixed by hand
- One-off SQL scripts repair inconsistencies: the schema no longer matches the business.
- The prototype found customers
- Built fast, often with AI, it proves demand but cannot carry real customers: authentication, payments, permissions, errors.
- The stack is no longer maintained
- An abandoned framework, dependencies with known vulnerabilities, hosting that no longer updates: the cost of staying grows every month.
02 — Method
How I run a rebuild
First the inventory of what the product really does — not what the documentation claims: the flows in use, the rules applied, the data and its inconsistencies, the integrations. I verify current behaviour before proposing to change it.
Then the model: a Postgres schema laid down from the business, the rules extracted into pure modules, tested against real cases and versioned where a past result must stay reproducible. That is what the old system lacked, and the reason for the rebuild.
Then the rebuild on Next.js — authentication, roles, payments, email, administration, observability — with a demo every week. Data migration is a repeatable script, dry-run several times before the cut-over; the old and new systems run side by side until the numbers match.
Finally the cut-over and hand-off: redirects, checks on the critical journeys, code and accounts in your name, architecture and operations documentation.
03 — Alternatives
What is not a rebuild
- A specific bug
- If the product holds but one flow breaks, a repair will do: Code Rescue, fixed price.
- A dated interface
- If only the look has aged, a redesign will do: Design Rescue (in French). A product rebuild does not change a theme.
- An imminent launch
- If the question is “is it ready for real users?”, it is a pre-launch check: Launch Rescue.
- A fuzzy scope
- If nobody knows yet what to keep, the architecture diagnostic produces the rebuild’s scope and price.
04 — Cost
What a rebuild costs
Usually CHF 25’000–80’000, fixed fee, from CHF 20’000. What moves the price: the number of entities and rules to carry over, the volume and state of the data to migrate, the integrations, the number of roles. Beyond CHF 80’000, the rebuild is split into phases, each fixed-fee and each delivering a usable product.
The price is set after scoping and does not move. When the scope is not clear yet, the paid diagnostic — CHF 2’400, fully credited against the rebuild within 60 days — produces the inventory, the target model and the price.
05 — Hire?
Should you hire instead?
A rebuild has an end; evolving the product does not. If you have a team, it can run the rebuild — provided it has the time to do it without stopping everything else. If you do not, hiring for a rebuild means employing someone who discovers the system while rebuilding it.
The common pattern: the rebuild as a fixed-fee project, built to be taken over, then the hire of the person who will evolve the product. The hire-or-outsource guide prices both options with official sources.
Is this the right service?
This makes sense if
- Your product has real customers or users, and its architecture slows every change.
- The business rules are scattered, untested, or known to one person.
- A prototype — built fast or with AI — has to become a production product.
- You want a fixed price, a cut-over without downtime, and to leave with the code and the accounts.
It is not the right service if
- You only want to change the look: that is a redesign.
- The product is a native mobile app, or an ERP to configure.
- The budget is below CHF 20’000: a targeted repair fits better.
- You are looking for ongoing hourly maintenance.
Proof — products I built and run
My own products, with their actual state. No client project is published yet.
- atamatax.comIn production · payments liveAtamatax — US expat tax infrastructureA versioned rules engine whose every conclusion cites its rule: a past result stays reproducible after a rule changes — what a rebuild must guarantee.Next.jsTypeScriptPostgresStripeVercelRead the case study →
- vellaci.chIn production · subscriptions liveVellaci — EU Cyber Resilience Act compliance for product manufacturersImmutable revisions and an exportable evidence vault: the data model of a regulated product, laid down before the interface.Next.jsTypeScriptPostgresStripeCycloneDX · SPDXOSV · KEV · EPSSRead the case study →
- horvera.comIn production · payments liveHorvera — flight compensation infrastructureA versioned eligibility engine that delivers every decision with its trace and sources, and pages generated from data.Next.jsTypeScriptPostgresStripePostHogRead the case study →
How it works
Four steps, fixed scope, price agreed upfront.
- 01
Diagnose
20 minutes, free. Business model, workflows, risks and product scope. If I’m the wrong person, I say so.
- 02
Architect and propose
Within 48h: data model, system boundaries, integrations, UX — and one fixed fee with dated milestones. No “starting at”.
- 03
Build
4 to 6 weeks of controlled iterations. A demo every Friday on a Vercel preview. You watch the system grow live — no black box.
- 04
Launch
Payments, analytics, deployment, operational workflows. You leave with the repo, the accounts, the pipeline — and 30 days of 24h replies.
- 05
Improve
Measured iteration on real usage: a fresh fixed-fee scope per increment, or nothing if your team takes over. No hidden dependency.
Frequently asked questions
How much does rebuilding a web application cost in Switzerland?
Usually CHF 25’000–80’000, fixed fee, from CHF 20’000. The price depends on the entities and rules to carry over, the data to migrate and the integrations. Beyond CHF 80’000, the rebuild is split into fixed-fee phases.
Should we rebuild or improve what exists?
Improve when the data model still matches the business and the problems are local. Rebuild when every change breaks something else, data gets fixed by hand or the stack is no longer maintained. The diagnostic decides with an inventory, not an impression.
How long does a rebuild take?
5–10 weeks depending on scope, with a demo every week. Larger rebuilds are split into phases.
How do you migrate data without downtime?
With a repeatable migration script, dry-run several times on a copy of the real data, and a period where the old and new systems run side by side until the numbers match. The cut-over is planned, reversible and checked on the critical journeys.
Can you take over a prototype built with AI?
Yes. I first verify what it really does, keep what holds — often the interface and the understanding of the need — and rebuild what does not: data model, authorisation, payments, errors. See Prototype or v1.
Have you rebuilt applications for clients before?
No client rebuild is published yet. The proof shown here is my own products — said plainly rather than dressing an in-house product up as a client reference.
Further reading
- Prototype or v1: the difference that decides your launchA prototype impresses in a demo. A v1 survives the first real customer. Confusing them is expensive.Building products5 minRead →
- Next.js architecture for a production SaaS: the decisions I make on every productWhat separates a Next.js project that holds from one that gets rewritten after a year comes down to about ten decisions. Here are mine, as applied on Atamatax, Horvera, Wabiro, SponsorsMetrics — and on this site.Next.js & full-stack9 minRead →
- What AI-generated SaaS founders should verify before launchAn app that renders is not a product ready to take money. Eight areas, one test per area, and the difference between declaring and verifying.SaaS engineering8 minRead →
Other services
- Architecture and product auditThe paid diagnostic before the build: v1 scope, target architecture, risks, plan and budget. CHF 2’400, seven working days, 100% credited against the build.
- SaaS v1 developmentThe first version that survives its first customer — what it must contain, and what can wait.
- Next.js developerProduction Next.js applications — App Router, Server Components, Postgres, Stripe — fast, indexable, maintainable.
A product that slows you down more than it moves?
Describe what it does, what breaks and what it costs today. Six questions to check the fit, a reply within 48h.
danilson.daveigaramos.pro@gmail.com · Reply within 24h, Mon–Fri.