Certificates are not proof that somebody can deliver a working AI system. They show that the individual has completed a structured body of learning and can be held against a defined knowledge baseline.
That distinction matters. AI has created a market full of impressive course lists, inflated capability claims, and systems that stop at the demonstration layer. The useful question is not whether somebody has collected a certificate. It is whether the learning is visible in the way they scope the problem, control the risks, structure the build, preserve human review, and define what the technology should not be allowed to do.
The three credentials shown here form a deliberate progression rather than a random training record.
The IBM Introduction to Artificial Intelligence course establishes the broader technical and business vocabulary. It covers the concepts behind artificial intelligence, machine learning, deep learning, generative AI, large language models, application areas, ethics, and responsible adoption.
The AI Academy AI Fundamentals series expands that foundation into the current working environment. It connects AI concepts with generative AI, prompting, AI agents, organisational adoption, and the effect these technologies have on individuals and businesses.
The OpenAI Codex Solutions Practitioner credential takes that learning into a more specific enterprise delivery context. It focuses on understanding how to position and scope Codex use in enterprise environments, rather than treating code generation as an uncontrolled shortcut.
Together, the credentials move from understanding what AI is, through understanding how it is used, to understanding how a specific AI development capability should be positioned within real delivery environments.
They sit on top of approximately 20 years of IT experience and an active body of AI product work. They do not create the capability. They validate, organise, and sharpen parts of it.
The OpenAI credential is particularly relevant to the way the Pl8ypus systems are being built. Effective use of Codex is not simply about asking an agent to produce code. It requires choosing the right problem, defining the boundaries, retaining evidence, preserving human review, and recognising when automation would introduce more risk than value.
The certificates support the case. The systems remain the proof.
