What sets IT expertise for software vendors apart
IT expertise for software vendors is aimed at companies for which software is not a business tool: it is the product they sell. SaaS vendors, platforms, fintechs, legaltechs: for them, an architecture decision ends up in the contract signed with the customer, in the margin and in the commercial roadmap. An outage does not merely inconvenience an internal department, it directly affects those who pay.
Three constraints recur from one company to the next. Pace first: the product evolves continuously, and every release must ship without degrading what works. Trust next: customers, often large organisations, ask for evidence of security, availability and compliance before signing, and again at each renewal. Dependencies finally: open source libraries, cloud services and AI models supplied by third parties make up a large part of the product, and the vendor answers for the whole to its customers.
The texts that apply to the product itself
For a software vendor, regulation does not only concern its internal data. It shapes the product and the commitments made to customers.
The Cyber Resilience Act. Regulation (EU) 2024/2847 sets cybersecurity requirements for products with digital elements, software included, and for their remote data processing solutions. Its Annex I requires the manufacturer, among other things, to draw up a software bill of materials, to address vulnerabilities without delay, including through security updates, and to subject its products to regular security testing. Its Article 14 requires notification of any actively exploited vulnerability, with an early warning no later than 24 hours after becoming aware of it. Under its Article 71, the regulation applies from 11 December 2027, with Article 14 applying since 11 September 2026 (text published in the Official Journal, read on 20/11/2026). Our entry on the Cyber Resilience Act presents its structure.
NIS2. Directive (EU) 2022/2555 covers in particular, in its Annexes I and II, cloud computing service providers, managed service providers and providers of online marketplaces, subject to the size criteria and exceptions of its Article 2. The risk management measures of its Article 21 cover supply chain security and security in the development and maintenance of systems, including vulnerability handling and disclosure. A vendor outside the scope of the NIS2 directive still feels its effects: its customers that are subject to it must take the security of their suppliers into account, and may require it in their contracts.
The GDPR. When it processes personal data on behalf of its customers, the vendor most often acts as a processor within the meaning of Article 4, point 8, of Regulation (EU) 2016/679, subject to its Article 28. Its customers, as controllers, also expect the product to help them comply with Article 25, on data protection by design and by default.
The AI Act. Regulation (EU) 2024/1689, as amended by Regulation (EU) 2026/1744, distinguishes the provider, which places an AI system on the market or puts it into service under its own name or trademark, from the deployer, which uses it under its own authority. A vendor that adds an AI feature to its product may become a provider, and its customer a deployer. Obligations then depend on the use: our entry on the AI Act explains the logic.
A fintech authorised as a payment institution, an electronic money institution or a crypto-asset service provider is also subject to Regulation (EU) 2022/2554, known as DORA, under its Article 2. A vendor that provides ICT services within the meaning of Article 3, point 21, to financial entities (an online service, for example) is also an ICT third-party service provider, a category referred to in Article 2, paragraph 1, point (u). Our page on banking and insurance details this framework.
Typical workstreams for a software vendor or platform
- Industrialising the delivery pipeline. Continuous integration and deployment, reproducible environments, progressive feature activation with feature flipping tools such as Unleash or LaunchDarkly. The subject falls under DevOps and platform engineering.
- Controlling the software supply chain. Dependency inventory, a bill of materials in CycloneDX or SPDX format produced by tools such as Syft or Trivy, vulnerability tracking, a coordinated disclosure procedure. These practices meet part of the requirements of Annex I of the Cyber Resilience Act and Article 21 of NIS2.
- Keeping service commitments. Service level objectives, alerts based on user experience, incident management and customer communication: this is the domain of observability and SRE.
- Integrating generative AI into the product. Choice of model, evaluation of answer quality, control of inference cost, data segregation between customers. Our page on generative AI in the enterprise details these trade-offs.
- Paying down the debt of a product that grew fast. Breaking up a monolith, stable interface contracts, test coverage sufficient to change the code without fear.
Running a technical workstream without freezing the product
The product roadmap remains the priority. A migration or an architecture overhaul cannot generally freeze the changes promised to customers. The workstream is broken down into stages that go into production one after another and coexist with current releases, each one reversible.
The end customer is in the loop. Service level agreements, announced maintenance windows, release notes, security questionnaires: every change has a visible effect on customers who have their own constraints. A well-run project produces, as it goes, the evidence the sales team and the security team will need: test results, code reviews, bills of materials, incident log.
The indicators are those of the product. Deployment frequency, time to fix a vulnerability, change failure rate, time to restore service after an incident: these observable measures show whether a technical workstream genuinely improves the capacity to deliver. They are weighed with the product owner against feature requests.
IT expertise for software vendors: how our consultants work
Management Square works on a time and materials basis. Our consultants join your product, engineering or operations teams, work with your tools and your rituals, and bring their method and expertise; decisions remain yours. They can strengthen a team in a specific area (architecture, platform, quality, data, AI) or contribute to managing a cross-cutting workstream, such as preparing a product range for the requirements of the Cyber Resilience Act, the interpretation of which is a matter for your lawyers.
With its roots in project, programme and portfolio management advisory, the firm helps a software vendor keep three calendars aligned: that of the product, that of customers and that of regulation.
Our approach follows four stages: frame, prove, industrialise, operate. At a software vendor, each one aligns with the product’s release cycle. To tell us about your product and the team to strengthen, contact us.