In this article, Axel RUDLOFF (founder and President of Koragence, and a student at ESSEC Business School, Grande Ecole Program – Master in Management (MiM), 2025–2029) shares observations drawn from building and managing software engineering projects involving system integration, artificial intelligence (AI), DevOps and custom business applications. The article explains why a software project is not only a technical investment: it is also a managerial decision about productivity, risk, operating processes and capital allocation.
Introduction
Companies rarely decide to develop custom software when existing systems no longer support the way the business actually operates. An enterprise resource planning system (ERP), such as SAP, may need to exchange data with a customer relationship management system (CRM), supplier application programming interfaces (APIs), document-management tools or applications developed internally. Difficulties arise when these systems must share data reliably, enforce specific business rules and support critical processes without repeated manual entry.
At Koragence, the projects I supervise often involve synchronising ERPs, integrating partner APIs, automating document processing with AI, rebuilding business workflows, or implementing DevOps architectures. DevOps refers to the practices and tools used to automate, secure and monitor the development and operation of software. Depending on the business criticality of a project, the architecture may target 99.95% availability, or up to 99.99% when the infrastructure supports multi-region redundancy, load balancing and advanced incident-recovery mechanisms.
About Koragence
Koragence is a French digital-services company specialising in custom software, web engineering, and process automation. The company designs business applications, software-as-a-service (SaaS) platforms, internal tools, customer portals and technical integrations for companies, startups, associations and public organisations.
Koragence has delivered more than 20 projects for clients located in more than six countries and works with a network of more than 50 active collaborators and partner companies. This model makes it possible to assemble a team according to the specific needs of each assignment, including software development, user experience and user interface design (UX/UI), cloud infrastructure, cybersecurity, accessibility and data engineering.
Logo of Koragence.
Source: the company.
My experience as founder and President
As founder and President, I am responsible for turning a business problem into a project that can be delivered economically and technically. This includes business development, qualification of client needs, project scoping, pricing, team selection, contractual discussions, delivery governance and long-term client relationships.
I also remain directly involved in product and technical decisions. On a typical assignment, I help identify the critical workflow, define the minimum useful scope, choose which components should be standard and which should be custom, coordinate the specialists involved, monitor delivery and manage relationships with clients.
My main responsibilities
My work can be divided into five areas: identifying operational problems with measurable business consequences; translating those problems into functional and technical features; building the right project team; controlling scope, budget, quality and delivery risk; and ensuring that the software creates value after deployment rather than becoming an additional tool that employees do not use.
Required skills and knowledge
This role requires both technical and managerial skills. Technical knowledge is necessary to assess architecture, security, integrations, databases and infrastructure. Business knowledge is equally important because the best technical solution is not always the best investment. A founder must also understand margins, cash flow, pricing, negotiation, contractual risk and the opportunity cost of allocating a team to one project rather than another.
The most important soft skills are active listening, synthesis, communication and decision-making under uncertainty. Clients rarely describe their problem in technical terms. It is therefore necessary to distinguish symptoms from root causes, challenge assumptions with diplomacy, and explain trade-offs to both technical and non-technical stakeholders.
What I have learned
The main lesson is that the quality of the initial diagnosis, the clarity of responsibilities and the realism of the scope have a major impact on project cost and delivery.
The invisible cost of a fragmented information system
One of the most frequent problems is the repeated entry of the same information into several systems. In one project for an industrial group with several hundred employees, a team spent more than 1,600 hours per year re-entering information from supplier catalogues into an internal database. Beyond the labour cost, this process created entry errors, inconsistencies between reference systems and disputes caused by contradictory information.
The software project removed this break in the information flow. Supplier data were collected automatically through standardized AI document extraction, checked for consistency and integrated into the internal SAP system. The economic value came from reducing labour and disputes with suppliers.
Integration has become a central management issue
Suppliers increasingly expose APIs, ERPs provide connectors, CRMs publish webhooks and most business software can exchange data automatically. This creates opportunities, but every external connection also becomes a dependency. A partner API may change version, become unavailable or modify its behaviour. Reliable software therefore requires monitoring, logging, error recovery, security controls and ongoing maintenance.
In another project, Koragence developed a platform capable of supervising more than 15 industrial machines through supplier APIs while centralising more than 1,100 alerts and distributing notifications across several channels. The architecture was designed to scale to 100 machines without replacing the underlying technical model.
This type of project illustrates why integration is a strategic issue. The system must not only work on launch day; it must continue to work when the number of users, machines, documents or external dependencies increases. Maintenance is therefore part of the investment decision from the beginning.
Artificial intelligence changes the economics of software and automation
In the past, automating supplier catalogues, technical data sheets, invoices and other documents required highly standardised formats and rules written separately for each source. In practice, these standards were often incomplete or inconsistently applied. AI models can now interpret a much wider variety of documents, extract useful information and feed databases or business software. This makes some automations faster and less expensive to implement.
However, AI does not remove the need for architecture, control and human judgment. A 2026 study of more than 100,000 software developers found that autonomous coding agents produced very large gains in coding activity, but smaller gains at the level of completed projects and actual releases. The authors interpret this difference as evidence that human and organisational bottlenecks still limit final output. In other words, writing code faster does not automatically mean shipping reliable software faster.
This distinction matters for managers. AI can lower the marginal cost of implementation, but the investment remains rational only when the company has correctly defined the workflow, data model, responsibilities, controls and expected return. The decision is therefore not “AI or no AI”; it is how AI can be integrated into a dependable operating system.
Economic, financial and business concepts related to my founder experience
I present below three concepts that are directly connected to my work at Koragence: transaction costs and the make-or-buy decision, return on investment and payback period, and operating leverage through reusable software assets.
Transaction costs and the make-or-buy decision
A company can buy standard software, adapt an existing product, outsource a custom development or build internally. The licence price is only one part of the decision. Managers must also consider transaction costs: integration work, manual reconciliation, training, vendor coordination, contract management, switching costs and the risk created by dependence on a supplier.
Standard software is usually preferable when the process is common and the product already satisfies the essential requirements. Custom software becomes more rational when the workflow is strategically important, highly specific, poorly served by standard tools or expensive to operate manually. My role is to help clients compare these alternatives rather than assume that custom development is always the correct answer.
Return on investment and payback period
Return on investment (ROI) compares the economic gains generated by a project with its total cost. For software, the benefits may include labour hours saved, fewer errors, faster sales cycles, reduced downtime, better compliance or additional revenue. The cost must include not only development, but also hosting, maintenance, training and change management.
The payback period measures how long it takes for cumulative benefits to recover the initial investment. For example, if an automation costs €40,000 and creates €5,000 of measurable monthly savings, its simple payback period is eight months. This calculation is not sufficient on its own, but it creates a common language between operational teams, finance teams and technical providers.
Operating leverage and reusable assets
Software can create operating leverage because the same technical system can support a higher volume of transactions without a proportional increase in labour. A platform designed for 15 machines and capable of supporting 100 machines illustrates this principle: the client can grow while avoiding the need to multiply manual supervision at the same rate.
The same logic applies to Koragence. Reusable components, documented deployment processes, quality controls and specialist partnerships reduce the cost and risk of future projects. Nevertheless, reuse must not become generic copy-and-paste delivery. The objective is to standardise the reliable foundations while preserving the business-specific layer that creates value for each client.
Why should I be interested in this post?
This topic is relevant to students interested in entrepreneurship, corporate finance, consulting, operations, private equity or digital transformation. Software investment decisions increasingly affect company valuation, operating margins, scalability and risk. Understanding these projects therefore requires more than technical knowledge.
For a finance student, custom software provides a concrete example of capital allocation: management commits resources today in exchange for expected future cash flows, cost savings or strategic flexibility. For a future consultant or entrepreneur, the article also shows why technology projects must be framed around measurable business outcomes rather than features alone.
Related posts on the SimTrade blog
▶ All posts about professional experiences
▶ Marco SIMONETTI Cristoforo Travel — From Zero to Exit: My Founder Story
▶ Alessandro MARRAS, Venture Capital 101: A Quick Overview
Useful resources
Academic research
Brynjolfsson, E., Li, D., & Raymond, L. R. (2025). Generative AI at Work , The Quarterly Journal of Economics, 140(2), 889–942.
Demirer, M., Musolff, L., & Yang, L. (2026). Writing Code vs. Shipping Code: Productivity Effects Across Generations of AI Coding Tools , NBER Working Paper No. 35275.
Williamson, O. E. (1989). Transaction Cost Economics , in Handbook of Industrial Organization, Volume 1, 135–182.
Lacity, M. C., Khan, S. A., & Willcocks, L. P. (2016). The role of Transaction Cost Economics in Information Technology Outsourcing research: A meta-analysis of the choice of contract type , The Journal of Strategic Information Systems, 25(1), 32–48.
Koller, T., Goedhart, M., & Wessels, D. (2020). Valuation: Measuring and Managing the Value of Companies, Seventh Edition, Hoboken (NJ), John Wiley & Sons.
Business resources
Koragence — When custom software still makes sense in 2026
National Bureau of Economic Research — AI coding tools and software delivery productivity
About the author
The article was written in July 2026 by Axel RUDLOFF, founder and President of Koragence and a student at ESSEC Business School, Grande Ecole Program – Master in Management (MiM), 2025–2029.
▶ Discover all articles by Axel RUDLOFF.
