Enter GenAI - a tool that I initially approached with some caution. Was it ‘cheating’? And could it really help us tackle the nuanced demands of competency framework design? I wasn’t looking for AI to replace the insights and expertise of our SMEs, but rather to augment our efforts effectively and efficiently.  

Partnering with Opencast Senior Solutions Architect Vipul Mogre, we explored together how AI might help us to streamline the framework development process. The collaboration with Vipul introduced me to the concept of ‘prompt engineering’ for large-language models (LLMs), and I quickly saw the potential of AI to speed up our work while preserving the human element. 

Building a competency AI: a unique framework Using a specific competency-focused LLM, Vipul and I began honing prompts to teach the AI about Opencast’s story, structure, organisational values, tone of voice and framework needs.  

I trained the AI model in the knowledge, methods, and tools (KMT) format, which is essential for each framework. By iterating the frameworks through the AI, I could then generate drafts that were more aligned with our organisational language, values and technical requirements with each round of feedback.  

Working with subject matter experts (SMEs), we produced input documents specifying key themes, focus areas and KMT distinctions. The AI then generated the output - comprehensive frameworks levelled for professional growth, ready to be reviewed and finalised by the functional SME.