Skip to content
MAWT Logo
Software Development

Data you can trust, queries that stay fast.

Overview

Schema design, query optimization, zero-downtime migrations and scaling on PostgreSQL, SQL Server and MySQL — by engineers who treat your data as what it is: the heart of your business.

Database development covers designing, optimizing and evolving the databases behind your software: schema design, query optimization, migrations without downtime, scaling and backup strategies. Done properly on PostgreSQL, SQL Server or MySQL, it determines whether your application stays fast and your data stays intact as the business grows.

Target

For Swiss SMEs whose applications have grown faster than their database: software slowing down, reports timing out, migrations everyone fears, backups nobody has ever tested. Also for teams starting a new product who want the data model right from day one.

Details

Every application stands on its database. When the schema is sound, everything else gets easier: features ship faster, reports run instantly, migrations stop being feared events. When it isn't, every sprint pays the tax.

We design, optimize and evolve databases on PostgreSQL, SQL Server and MySQL — schema design, query tuning, zero-downtime migrations, scaling and tested backups — for brand-new products and for systems that have been running for fifteen years.

Why the database quietly decides your software's future

Every application stands on its database. When the schema is sound, everything downstream gets easier: features ship faster, reports run instantly, and changes stop being scary. When it isn't, every new feature pays a tax — slow queries patched with more hardware, workarounds stacked on workarounds.

Schema decisions made in week one are still shaping performance ten years later. That is why we put senior engineers on data modelling, whether we are designing a new product's database or untangling one that has grown organically since 2010.

Fixing slow: how we optimize an existing database

We never guess. Optimization starts with measurement: query plans, slow-query logs, lock contention, real production load. The findings usually rank clearly — a handful of queries and missing indexes typically cause most of the pain.

Then we fix in order of impact: indexes and query rewrites first, then structural changes where the model itself is the bottleneck. You get before-and-after numbers for every intervention, in writing — not an invoice and a promise.

Migrations without holding your breath

The word migration makes teams nervous because they have seen one go wrong. It doesn't have to be that way. We engineer migrations as code: versioned, tested against a copy of production, reversible, and executed with zero-downtime techniques so the business never notices.

This applies to schema changes, engine upgrades, and moves between systems — SQL Server to PostgreSQL, on-premises to a Swiss or European cloud — including for applications built by someone else.

Backups that actually restore

A backup that has never been restored is a hope, not a backup. We set up backup strategies with automated restore tests, defined recovery objectives, and integrity checks that catch silent corruption before it spreads.

For data covered by the nLPD, we also make sure of where the data — and its backups — physically live, with Swiss or European hosting when your obligations require it.

What it includes
  • Schema design and data modelling that match your business
  • Query optimization and indexing on PostgreSQL, SQL Server and MySQL
  • Zero-downtime migrations, versioned and reversible
  • Scaling strategies: partitioning, replication, caching
  • Backup, restore testing and data-integrity guarantees
  • Audits of existing databases with a prioritized action plan
Deliverables
  • A documented data model aligned with your business processes
  • Measured performance gains — query times before and after, in writing
  • Versioned, reversible migrations executed without service interruption
  • A tested backup and restore procedure, not just a scheduled job
  • A prioritized roadmap for scaling as your data grows
Takeaways
  • Schema decisions made early shape performance for years.

  • Most slow applications are slow databases — and most are fixable with measurement, not hardware.

  • Migrations can run without downtime when engineered, not improvised.

  • A backup that was never restored is a hope, not a backup.

  • PostgreSQL, SQL Server, MySQL — senior expertise, in Geneva.

You might also need

MVP Development

MVP development for startups and founders: a working product in a few weeks, ready to meet the market.

AI Development

Custom AI development for Swiss companies: AI agents, fine-tuned models, ML pipelines and LLM integration (Claude, ChatGPT, Gemini) — from prototype to production.

Web Development

Custom websites for SMBs and growing companies. Design, development, launch, iteration.

Frequent questions

Our application keeps getting slower. Is it the database?

Very often, yes — and usually a small number of queries and missing indexes cause most of it. A short audit measures where the time actually goes and produces a prioritized fix list. Many performance problems are resolved in days once they are measured instead of guessed at.

PostgreSQL, SQL Server or MySQL — which should we use?

For new projects, PostgreSQL is our usual recommendation: powerful, open source, no licence costs. SQL Server makes sense in Microsoft-centric environments, and MySQL remains solid for many workloads. If you already run one of them, the right answer is usually to master it rather than migrate — we advise per case, without dogma.

Can you migrate our database without interrupting the business?

Yes, in almost all cases. We use zero-downtime techniques: the migration is rehearsed on a copy of production, versioned, reversible, and switched over in a controlled window. The engineering effort goes into making the day itself boring.

Do you work on databases you didn't build?

Constantly — it is most of our database work. We audit the existing schema, queries and infrastructure, document what we find, and improve it step by step. No judgement about past choices, no pressure to rewrite: we work with what exists.

How do you guarantee our data's integrity and backups?

With constraints and transactions enforced at the database level, automated integrity checks, and — crucially — restore tests that actually run. We define recovery objectives with you (how much data loss and downtime is acceptable) and build the backup strategy to meet them, then prove it regularly.

Where will our data be hosted?

Wherever your obligations require: your own servers, a Swiss data centre, or a European cloud. For nLPD-sensitive data we document the full picture — primary storage, replicas and backups included — so you can answer an auditor's question with a diagram, not a shrug.
Next steps

Your data is the business. Treat it that way.