Methodology
How we decide what to recommend
Every plan EduVentureSpace generates is built from three sources, combined with the help of large language models that we keep on a tight leash — they explain and connect the data, they don't invent it.
The three inputs
- 1. Official university data
- Entry requirements, programme details, application deadlines, and English-language thresholds — pulled from each institution's own pages, not from third-party rankings or aggregators. When a fact comes from a non-official URL, we mark it as researched rather than verified, and the path itself shows the difference.
- 2. Public outcome data
- UK applications and offers data from UCAS, post-graduation earnings from the ONS Longitudinal Education Outcomes dataset, and equivalent public sources where they exist for other countries. We do not buy proprietary datasets that we couldn't cite back to the original publisher.
- 3. Your own answers
- Current grades, country, target subjects, scholarship needs, and the constraints that actually matter to you. You can update any of these later; the path regenerates without losing your previous version.
Where we stop short
We don't recommend programmes we can't verify against an official source. When the evidence is approximate — older entry-requirement pages, programmes still being renamed, scholarships whose 2026 numbers haven't been published — we surface that uncertainty on the path itself, with a confidence label, instead of rounding it into a clean number.
The aim is to be useful to a sixteen-year-old making a real decision, which means being upfront about what we know, what we're inferring, and what is still open.