The companyContext field: personalising learning paths for your stack

LearningPathInput has an optional companyContext field: a string of up to 500 characters describing the learner's team, stack, and context. You can pass it today - the schema accepts it and validates it. It does not yet change the generated path. That is intentional, and this post explains why.

We shipped the field before the feature for two reasons. First, it locks in the API shape before any consuming code depends on a version without it. If we added it in v1, every product that had already shipped LearnKit AI would need to handle a schema change. By shipping it in v0, all valid v0 inputs are valid v1 inputs - the only change is that the engine starts doing something with a field it was already accepting.

Second, we needed to decide what "using companyContext" actually means before we built it. The obvious design is string interpolation: embed the context into lesson summary templates wherever there is a gap. The better design is semantic: parse the context for signals (team size, stack, industry, pace) and use those signals to weight lesson selection, adjust project scope, and tune the language of summaries.

In v1, companyContext will be parsed for four signals: tech stack keywords (to prefer lessons that use tools the team already has), pace signals ("ships weekly" versus "quarterly releases"), industry signals (regulated industries get different compliance framing), and team size signals ("40-person team" versus "solo founder"). The lesson templates will expose hooks for each signal so the interpolation is structured, not arbitrary.

If you are building on LearnKit AI today, pass companyContext. It will be silently ignored for now. In v1, it will start doing something useful - and your integration will not need to change.