Frédéric Delchambre 25+ years of experience · Business applications · Architecture · Data · AI · Security

Let's not build your idea just yet

For over 25 years, I have focused on identifying the real problem first, then designing the simplest, most robust and useful solution to address it.

Before choosing a technology or starting development, I put the request back into the real context of your business: objectives, processes, people, data, tools and constraints.

I go into exactly as much detail as needed — no more, no less — to quickly reveal the dependencies, risks and decisions that matter.

I can then design and develop the solution, define how it should work and its architecture with your teams or partners, or lead its delivery through to production.

Let's talk for 30 minutes No requirements document needed.

01 Approach

Understand before building

An application never operates in isolation. It sits within a set of processes, people, data, habits, constraints and tools that have evolved over time.

Look beyond the application

I take an overall view of how things really work: the processes as they are intended to operate, but above all how they are actually used day to day.

Technical environment. ERP, business applications, APIs, Excel files, internal tools, databases, AI, infrastructure and suppliers.

Regulatory framework. GDPR, the AI Act, NIS2 and sector-specific rules where applicable — to be built in from the start, not added afterwards.

People involved. Users, decision-makers and the people who will operate, maintain and evolve the solution.

Change only what needs to change

A process that looks strange often exists for a reason. An old application may embody fifteen years of business rules and know-how; an Excel file may be perfectly suited to its role.

I first try to understand what works and why before proposing any change.

The goal is not to impose a new way of working, but to determine what is worth keeping, simplifying, connecting, automating, replacing or developing.

02 Situations

A few good reasons to call me

You do not need to have already defined the solution.

An important problem, an idea worth challenging or a situation that has become difficult to control is enough to get started.

Your tools and applications are becoming difficult to manage

ERP systems, business applications, Excel files, interfaces, automations and manual processes have accumulated over time.

Everything still works, but it is becoming difficult to get a clear view of the dependencies, risks and what should be improved first.

You are considering a new application or a major evolution

The need is real, but you want to verify the best way to address it before committing time, budget and resources.

A new application, replacing a tool, automation or integration: the idea deserves to be challenged before it is developed.

You need someone to lead a project and coordinate everyone involved

The need is defined, but delivering it requires technical expertise and involves several teams, service providers or suppliers, with dependencies, a schedule and a budget to manage.

I take care of project leadership: planning, coordination, budget and dependency tracking, and priority decisions. My technical background enables me to challenge proposals and anticipate risks.

You need someone to take over an existing application or project

The system works, but understanding of it has deteriorated over time.

Documentation is incomplete, some decisions are no longer clear, technical debt is increasing or a significant part of the knowledge depends on only a few people.

The first objective is to regain a clear view before deciding what should change.

You want to automate or integrate AI without following the hype

AI and agentic systems open up new possibilities, but their value must be assessed within the real operation of the business.

We identify use cases where the impact can be measured, then examine what they mean in production: APIs, tokens, infrastructure, monitoring, security and costs.

The goal is not to add AI simply because the technology allows it or because it is fashionable, but to use it when it makes the solution genuinely more efficient or profitable.

Your new service or product needs an application

Customer portal, B2B service, internal tool, smart kiosk or an idea that is still loosely defined: we examine how the application should work, how it will integrate with your existing systems and what actually needs to be built so that it remains reliable and maintainable over time.

03 Challenge

The obvious answer is not always the best solution

Challenging an idea does not mean slowing it down.

It means checking that the proposed solution actually addresses the real problem before developing, replacing or automating anything.

Case01

“We need a new application for our sales teams”

The request sounds straightforward.

But when we look at day-to-day operations, we discover that salespeople enter the same information several times, maintain their own Excel files and manually perform overlapping tasks.

Adding another application could simply add another layer.

The priority may instead be to consolidate the data and its flows before deciding what is actually worth developing.

Case02

“We want to use AI to automate part of our customer service”

The idea may be relevant.

Before choosing a model or building an agent, we first need to understand what generates the requests, which data will be used and whether AI is addressing the right need — rather than merely the symptoms of a process that could be improved.

We then need to decide where and how the AI will run.

An external API is suitable for some use cases. In others, confidentiality, volumes or recurring costs may justify running a model locally — while still accounting for infrastructure and maintenance costs.

The goal is to choose an architecture that delivers measurable value while retaining control over data, risks and costs.

Case03

“This old application needs to be replaced”

Maybe.

It is ageing and its interface is no longer modern.

But it often contains years of business rules and decisions that are documented nowhere else.

Replacing it without understanding those rules can lead to rebuilding an application that is more modern… but less suited to the business.

The first task is therefore to reconstruct that knowledge, identify what is worth preserving, and then decide how to replace or evolve the application.

04 Engagement

Take stock before moving forward

Before investing, let's spend a few days getting a clear view.

In just a few days, we can establish the context, verify the priorities and avoid committing too quickly in the wrong direction.

This short engagement is designed to understand the situation, identify what deserves your attention and determine the next steps before committing significant budget or resources.

The stages of the engagement

01
Understand
I talk with the people involved — management, business teams, IT/ICT, users or partners — to understand the need, the existing environment and the constraints.
02
Map
Applications, data, flows, processes, dependencies and key stakeholders: the objective is to quickly obtain a sufficiently clear view of the situation to make informed decisions.
03
Identify
I highlight friction points, technical, operational or regulatory risks, wasted time, dependencies and improvement opportunities that deserve attention.
04
Challenge
We test the proposed solutions against reality: expected value, constraints, impacts, risks and level of investment.
05
Prioritise
We determine what is worth doing now, what can wait and what is better left undone.

Deliverable

What you get from it

The result should be useful, not lengthy for the sake of it.

You leave with a structured summary: a map of the existing environment, the main findings, identified risks and opportunities, recommendations and priorities.

This summary should make it clear:

what is worth keeping;
what should be improved or simplified;
what can be automated or connected;
what requires development;
what probably does not justify the investment;
and in what order to proceed.

The objective: invest a few days to make better decisions before committing to a full project. The next phase can then be delivered by me, your teams or a partner, on a clear and shared basis.

05 Projects

Real projects, built to last

More than 25 years designing, developing and evolving business applications.

My approach is backed by more than 25 years of experience and by projects that have genuinely been put into production across a wide range of environments and industries.

I continue to evolve several of these solutions as the needs of their users and the businesses operating them change.

Online B2B sales across the Benelux

INTERNATIONAL B2B PLATFORM

25,000+ products · thousands of B2B users · BE / NL / LU

B2B e-commerce platform for a major electronics retailer, used by dozens of corporate customers, each with its own purchasing rules, product selection and pricing.

It is integrated with supplier feeds (stock, pricing, catalogue and leasing), ordering systems and third-party HR platforms giving employees access to cafeteria plans.

In production for more than five years, it has seen strong growth since its expansion into the Netherlands and Luxembourg.

Full lifecycle management for sales agents

END-TO-END MANAGEMENT PLATFORM

Target: more than 2,000 agents by early 2027

Platform developed for Belgium’s second-largest electricity supplier, enabling its partners to manage the complete journey of their sales agents: onboarding, examination and certification, appointment scheduling, customer service forms, reporting and Salesforce integration.

This process was previously managed manually using Outlook and Excel.

In September 2026, around 600 agents are active; that number is expected to exceed 2,000 in early 2027 with a new partnership. The platform has been continuously evolving for more than five years, including an internal AI assistant currently being designed.

Training and tracking platform for drone pilots

BELGIAN DEFENCE

Several hundred military personnel ultimately involved

Design of a training, assessment and tracking platform for Belgian military personnel operating drones: identity verification, online examinations with video monitoring and full training-path tracking. The project is now expanding with a second platform dedicated to Specific category training, including EU-STS, alongside Open A1/A3 and Open A2 certificates, while incorporating the European EASA framework and Belgian-specific requirements from DGTA and Skeyes.

Knowledge base for sales teams

AI AND KNOWLEDGE BASE

From scattered information to a business chatbot

Previously, each salesperson managed their own documentation across emails, a file server and local folders. I developed a natural-language chatbot connected to a product database, FAQ and use cases; it can answer questions such as: “Which products fit this need?” The knowledge base grows with new field feedback and progressively captures the expertise of the whole team.

Large-scale e-shops for established brands

B2C E-COMMERCE

Tens of thousands of customers, hundreds of thousands of transactions

Development of numerous B2C e-commerce platforms connected to ERP systems and industry-specific data providers, including one for a major Belgian bookstore located in the heart of Brussels, offering more than 2 million books online.

I have also designed and developed e-commerce platforms for companies operating across a wide range of sectors, including machinery and tools, construction materials, hardware, supplies for garage networks, furniture and interior design for the international hospitality industry, parapharmacy, sports equipment, professional equipment for the meat industry across the Benelux and France.

Together, these platforms serve tens of thousands of customers, process a significant volume of orders, and generate several hundred thousand euros in revenue every month.

Headless CMS and API for custom web interfaces

BACK-END AND API

57 endpoints exposed by the API

Design of a headless CMS and its API, used by dozens of websites. Development of a shared payment gateway connected to Worldline, Mollie, Stripe and MultiSafepay, processing dozens of transactions every day. The CMS also serves as a foundation for third-party teams: several web interfaces are developed by partner agencies and hosted on modern platforms while relying on this common back end.

Custom interactive kiosks

KIOSKS · RASPBERRY PI

Several dozen retail locations

Design of interactive kiosks deployed across dozens of retail locations.

Based on Raspberry Pi and Debian Linux, they synchronise their data and transmit readings in real time. Custom adapters were designed, modelled and then 3D-printed to integrate the enclosures into the kiosk furniture. Video tutorials were also produced using Meta glasses.

Penetration testing and hardening of a B2B application

APPLICATION SECURITY

Remediation required by a major bank

Security assessment of a B2B platform combining e-commerce and logistics management. The main issues identified involved authorisation controls, SQL injection risks and insufficient validation of certain inputs. Correcting them enabled the platform to meet the security requirements imposed by one of its main end customers, a major Belgian bank.

Platform for a local television broadcaster

MEDIA AND CONTENT

45,000+ articles · 35,000+ videos · 80,000+ page views per day

Platform developed for a local television broadcaster, including the public website, editorial back office and a custom video player with pre-roll advertising management. More than 45,000 articles and 35,000 videos are available, with an average of more than 80,000 page views per day. The platform has been in production for many years.

Technology choices

Technology choices driven by the need

Understand new possibilities without imposing them.

Application architecture, web development, APIs, databases, AI, security, infrastructure, embedded systems and connected hardware…

I stay close to the practical side: depending on the problem, I can move from a business workshop to an architecture, a database, an API, code or a hardware prototype.

I keep up with technology continuously to understand what new approaches can bring, without trying to impose them when a simpler or more proven solution is preferable.

I choose and combine technologies according to the need and its context: constraints, existing systems, security requirements, the applicable regulatory framework and available skills, rather than starting with a solution that has already been decided.

When I build an application, I can also take care of its deployment, hosting, monitoring and ongoing evolution.

06 Involvement

Involved from start to finish

A clear line of continuity, from problem to solution.

You work directly with the person who understands the need, analyses the constraints, proposes the options and, depending on the project, designs or delivers the solution.

This continuity limits the loss of understanding between business teams, analysis, architecture and what is ultimately delivered.

When I develop the solution, I remain involved in what is actually built: application, data, integrations, deployment, infrastructure or connected hardware depending on the project.

If delivery is handled by your IT/ICT team, another team or an external partner, I can also define the functionality and architecture, transfer the context and support the implementation.

One point of contact does not mean working alone

I collaborate with the people already in place: management, business teams, IT/ICT, developers, suppliers or agencies.

What about dependency on a single person?

It is a valid concern, and one I prefer to address from the design stage: an application should not become difficult to take over simply because its lead developer is no longer available.

I favour understandable architecture, consistent conventions, readable code and documentation maintained throughout the project.

Important decisions remain explicit: how the solution works, why certain choices were made, where the dependencies are and how to operate it.

A well-structured solution is much easier for another team to take over. AI tools can accelerate that handover, provided the project was designed properly from the outset.

The goal is not to pretend that all dependency on an individual can be eliminated, but to make sure it never becomes an obstacle to the continuity of the project.

07 Contact

Let's clarify your project

Shall we talk for 30 minutes?

No requirements document needed. You do not need to have already defined the solution either.

Tell me in your own words what you want to improve, the problem you are facing or the idea you are considering. It is my job to adapt to your context, not yours to translate your needs into technical language.

In 30 minutes, we can usually identify the heart of the problem, spot a few areas that need attention and outline some initial directions.

It is also an opportunity to see whether my approach suits you, whether we share the same understanding and whether working together makes sense.

Half an hour is often enough to determine whether it is worth going further.

No unnecessary form.
Direct, simple and practical contact.

Let's talk for 30 minutes

Email · WhatsApp · Phone · LinkedIn

Emailfred@nomadly.be WhatsAppSend a message Phone+32 478 55 07 19 LinkedInView profile