Services · EN / DE · US-based
The right stack for the job, with proof attached.
Web, mobile, desktop, or a legacy system that needs a future: TypeScript, C#, Swift, Kotlin, SQL, or Delphi, whichever your problem calls for. Seven things I do for customers, each one linked to a system I shipped or an article that shows the work. No method vocabulary, no promises I cannot demonstrate.
01 — Service
Custom business systems
Warehouse management, B2B web shops, delivery apps, reporting pipelines — built from scratch in TypeScript, C#, or whatever your environment already runs, wired into your ERP, and shipped to production.
The usual starting point
Off-the-shelf software covers eighty percent of the process and the remaining twenty percent is where the company makes its money.
What you get
- Web applications built around your actual workflow: picking, packing, receiving, ordering, dispatch
- Integrations over OAuth 2.0 and REST with the ERP, the shop, the label printer, the PDF service
- Role-based access, passkeys and two-factor sign-in, audit-friendly history
- Automated releases with schema migrations, health checks, and a documented rollback
How
The stack is a decision, not a habit: Next.js and PostgreSQL where a browser app is right, .NET where your team lives in C#, Swift or Kotlin where it has to be native. One codebase per system, a contract document for every API between systems, and no caching that hides stale stock. You get the repository, the release scripts, and release notes in your language.
Proof
02 — Service
Apps for every platform
Web first, then native where it counts: iPhone, iPad, Apple TV, and Mac in Swift, Android in Kotlin, Chrome extensions — one backend, one set of rules, every screen your users actually hold.
The usual starting point
The web app works, but the field crew wants it on the phone, the boss wants it on the TV in the lobby, and a second team for every platform is not in the budget.
What you get
- A web application that is the single source of truth for data, rules, and sign-in
- Native apps that share that backend and its passkey sign-in instead of reinventing it
- Offline-capable mobile apps that keep working without a signal and sync when it returns
- Browser extensions and desktop shells where the workflow lives in another program
How
I design the system once and let the platform follow. The native apps are thin on purpose; every rule that matters lives on the server where it can be tested once. AI tooling makes the platform-specific code cheap to produce — I make sure it is also correct.
Proof
03 — Service
AI-assisted engineering, verified
Spec-driven development with Claude Code in any language, local LLMs embedded in your product with no data leaving the building, and independent review of everything the model wrote.
The usual starting point
Everyone is promising ten-times productivity; nobody is showing which parts of the output they actually checked.
What you get
- Working features built from written specifications, with the AI session driven by project rules I have refined over months
- Vision and language models integrated into your product — label scanning, document extraction, assistants — cloud or fully local
- Code review of AI-written pull requests, with the findings and the fixes
- Training for your team on the same workflow, so the leverage stays in-house
How
The model is a strong junior engineer with no memory. I supply the memory: architecture, conventions, framework versions, the legacy sources as read-only reference, and a test before every merge. The language is whatever fits — TypeScript, C#, Swift, Kotlin, Delphi — because what I bring is 35 years of knowing what correct looks like, not a favorite syntax.
Proof
04 — Service
Legacy to web migration
Delphi VCL, multi-tier Delphi, C# services, Firebird, InterBase, MS SQL — rebuilt on Next.js, React, and TypeScript with the data moved over and the business rules kept intact.
The usual starting point
The application still runs the business, but it is Win32-only, the database engine is out of support, and every new hire asks why there is no browser version.
What you get
- A complete Next.js application with a new relational schema, authentication, and role-based access
- A data migration tool with a progress UI that moves the real production rows — not a one-off script
- Charts, reports, PDF and Excel exports recreated feature for feature, then improved where the old app had stopped
- Docker images, deployment scripts, backups, and rollback so the customer owns the result
How
I read the old units, the service layer, and the schema first, because a migration that ignores twenty years of business rules produces a pretty app that computes the wrong numbers. Three decades of Delphi mean I can read code most web developers cannot; the new code is written with AI tooling under my project rules and verified against the old behavior before anything ships.
Proof
05 — Service
Delphi consulting and code rescue
Architecture reviews, FireDAC and database modernization, REST and JSON APIs, Docker and PostgreSQL for Delphi shops — three decades of Delphi on call, and a way forward when you want one.
The usual starting point
The original developer left, the component vendor folded, or the app needs to talk to a system that speaks JSON and it does not.
What you get
- A written review of the codebase with a prioritized list of what to fix, what to leave, and why
- Hands-on work in the code: FireDAC migrations, RTTI-based serialization, TMS XData and WEB Core services
- A reproducible development environment with Docker, PostgreSQL, and version control
- Compatibility reviews of backend changes against the clients that depend on them
How
I compile it, run it, and read it before I say anything. Findings are split into what the customer can change and what they should not touch. If the right answer is to keep the Delphi backend and put a modern front end on it, I say that too.
Proof
06 — Service
Training and courses
Live training for teams, video courses, and five books — AI-assisted engineering, Next.js and TypeScript, Delphi, multi-tier architecture, and the road from desktop to web — in English and German.
The usual starting point
The team knows one stack cold and has never shipped on another, or has the AI tools and no process that makes their output trustworthy.
What you get
- On-site or remote workshops built around your codebase, not slides about somebody else's
- Self-paced video courses on Delphi, TMS components, and Next.js
- Books that walk through a complete application from data model to deployment
How
Every example compiles, every project in the courses is one I ship in production, and questions after the training are answered.
Proof
07 — Service
Long-term partnership
An expert on retainer for teams that need one: US-based, Eastern time, fluent German for DACH customers, and the same person on the phone every time.
The usual starting point
You do not need a full-time developer, but you do need someone who knows the system and answers this week, not next quarter.
What you get
- A monthly block of hours with a named contact — me — and no hand-offs
- Ongoing releases, dependency upgrades, security patches, and backend reviews
- Availability in US Eastern and Central European business hours
How
Customers I work with this way have been with me for years. The relationship is the product.
Proof
08 — How I work
Four steps, no surprises.
/01
A conversation
Thirty to sixty minutes, no charge. You describe the system and the pain; I ask the questions a migration or a build will raise anyway. If I am not the right person, I say so and point you somewhere better.
/02
A fixed-scope assessment
For existing code: I read it, compile it, run it, and deliver a written review with a prioritized plan and a realistic estimate. You own the document whether or not we continue.
/03
Build, verify, ship
Work happens in a repository you have access to from day one, in short releases you can run. Every feature is verified against the old behavior or the written specification before it ships.
/04
Stay, if you want
Most customers keep me on afterward for releases, reviews, and the next feature. Retainers are monthly and cancelable.
Where I am
Based in Florida, fluent on both sides of the Atlantic.
FlixEngineering is a US company in Southwest Florida, and that is where I live and work. I am equally at home in English and German, so customers in the US get a local partner in their time zone, and customers in Germany, Austria, and Switzerland get contracts, meetings, and documentation in German without a translator in between.
- Company
- FlixEngineering, LLC
- Based in
- Estero, Florida, USA
- Working hours
- US Eastern and Central European
- Languages
- English and German, both fluent
- Stacks
- TypeScript, C#/.NET, Swift, Kotlin, SQL, Delphi
- In the business since
- 1991 · Embarcadero MVP since 2016
- Delivery
- Remote worldwide, on-site by arrangement
09 — Say hello
Bring the system. I'll bring the plan.
A first conversation costs nothing and usually saves a lot.