%20par%20Anthropic.webp)
The integration of LLMs and AI agents into software development is no longer experimental. Yet the growing set of tools — Claude Code, Gemini, DeepSeek, Qwen, Cursor, Ollama, Copilot and others — raises hard questions for tech and product leaders: How do you keep code secure? How do you prevent IP leaks? How do you build strong AI-powered systems?
To tackle these issues, Novity started a major effort in early 2026: certifying its teams through Anthropic's Claude Certified Architect -- Foundations (CCA-F) programme.
Continuous learning and product engineering as prerequisites
Before certification comes skills growth. Our approach is based on the need to always adapt our expertise to the realities of the tech market. At Novity, we do not see engineers as people who only execute predefined tasks.
The goal is not to teach engineers how to "prompt" an AI model. It is to reshape the value they bring to clients. With AI, faster code is now expected. The real challenge is knowing what to build and why.
Our engineers are tested on their ability to understand business problems, question functional needs and work well in complex enterprise settings. In other words, they adopt a Product Engineer mindset.
This is where artificial intelligence becomes a key boost — not a replacement for engineers, but a way to help them move faster.
Supporting this shift requires ongoing learning through our internal academy. By strengthening this culture of tech agility and AI fluency, we make sure our developers based in Mauritius and Madagascar can support our support clients' business-critical projects.
The Claude Certified Architect (CCA-F) Programme
Anthropic's Claude Certified Architect -- Foundations (CCA-F) certification is not just a test of theory. It is a hard check of an engineer's ability to design, integrate and run solutions built around the Claude family of models.
In practice, a CCA-F-certified engineer builds skills in four main areas:
- LLM system architecture: understanding how to continue a foundation model with vector databases to build systems such as Retrieval-Augmented Generation (RAG) architectures.
- Context and token management: how to use large context windows to process big codebases or dense docs while keeping inference costs under control.
- Security and compliance: how to set guardrails to reduce errors, protect sensitive data and make sure models stay within company rules.
- AI agent orchestration: how to move beyond simple text generation and design AI agents that can handle complex, step-by-step tasks.
Building sustainable AI governance beyond the hype
Investing in AI is not just a matter of giving a few employees access to new tools.
Rather than training a small group of isolated specialists, Novity has built a wider certification programme across its teams. AI adoption is a deeper company-wide challenge, and our approach is meant to address it as a whole.
As Xavier Lernould, CEO of Novity, said in a June 2026 LinkedIn post:
“Theory without code is just hot air. Code without understanding is just debt.”
Committing our teams to this programme represents an investment of nearly 80 hours per person. It requires time, hands-on work with real-world use cases and structured follow-up throughout the learning process.
By mid-2026, 42% of our teams had already started the Claude Certified Architect pathway, taking part in practical sessions and sharing feedback from real-world projects.
By autumn 2026, we are celebrating our first CCA-F-certified engineers, with 50% of our consultants currently moving through the learning programme.
The certification confirms their ability to integrate Anthropic models securely — expertise that is especially relevant when building bespoke SaaS solutions.
What Claude Certification means for our clients
Choosing Novity as a technology, product, data and AI partner means working with engineers who understand not only software development, but also AI system design, data security and the ability to scale.
Bringing Novity engineers with Anthropic Claude certification into your technology, product and data teams changes how projects can be handled:
- Greater design confidence: Your system design is not being handled by occasional users of the technology, but by engineers whose expertise has been validated through a certification programme created by Anthropic.
- Secure co-development: Our engineers support your projects within a clear method framework for managing and protecting data.
- Faster POC and MVP development: Whether you want to add generative AI features to an existing product or launch a new project, our teams know how to design the architecture from day one with scale in mind.
Sources
- BDM, Blog du Modérateur - Anthropic Academy: How to learn about AI and Claude
- Distributique - Anthropic strengthens its partner programme to support Claude adoption
- ChannelNews - Anthropic invests $100 million in its partner ecosystem
About Anthropic's CPAbout Anthropic's Claude Partner Network
The Claude Partner Network is how Anthropic works with the firms that put Claude into production for customers. The Services Track recognizes partners for building a certified delivery team, delivering Claude use cases for clients, and proving the work through public customer references. Partners advance through Certified, Select, Preferred, and Global Premier tiers.
