Picture two people who get the same invitation on the same day. One is a judge in a municipality where almost every domestic violence case goes through her court. The other works in the administrative area of a court in the capital and rarely sees a case file. The invitation says the same thing to both: a training cycle on gender and intersectionality, 14 courses and 400 hours.
Both have impossible schedules. Both will open the first course in good faith. And it is very likely that, by the third week, neither of them will feel it was designed with her in mind.
That was the risk in front of us when we started working with the Rodrigo Lara Bonilla Judicial School.
01 — The challengeTraining a whole country, without training everyone the same
Colombia’s judiciary is huge and diverse. There are high-court judges and trial judges, people who prepare and process cases, people who deal directly with victims and people who keep the administrative side running. All of them need to understand the gender perspective. None of them needs exactly the same thing.
The brief was to design a complete online training cycle, in three levels, on Moodle. The natural temptation in a project like this is to build the best possible course and open it to everyone. It is the easiest thing to manage, measure and report. It is also what works worst when the audience is this varied.
02 — The conversationAsk before you design
The decision that changed the project was to start with a diagnostic. Not a satisfaction survey, but an instrument to find out what each person knew and where the real gaps were.
The results were revealing, and not always in the direction we expected. Concrete confusions showed up around concepts everyone assumed were understood: what intersectionality really is, how to recognise psychological or economic violence, how to weigh evidence with a differential approach. And those confusions were not spread evenly: they depended on the role, the type of court and how close people were to victims.
Not everyone needed the same course. That finding, simple as it was, reorganised everything else.
03 — What we builtFrom diagnostic to path
With that data, the teaching team defined four learner profiles: those who decide, those in direct contact with victims, those who prepare and process cases, and those who support the work from the administrative side. Five paths were designed for them. All of them share a common core of three courses, because there is a foundation nobody should skip. From there, each path goes its own way. The fifth, meant for those who train others, covers all fourteen courses.
The technical challenge was that Moodle, on its own, doesn’t do this. It lets you organise courses, track progress and issue certificates, but it has no way to take a self-assessment, interpret it and turn it into a personalised path that the institution can follow.
So we built it. A Moodle module of our own, written from scratch, that does three things:
- Self-assessment. Each person completes an instrument that gauges their knowledge and their work context.
- Suggested path. From those answers, the tool places the person in a profile and proposes the level and courses they need.
- Follow-up. The School can see progress by profile, by path and by gap, not just how many people enrolled.
One technical decision mattered more than it might seem: the module lives inside the courses, as just another Moodle activity. That means it respects permissions, backups and the way the School already runs its platform. A tool that forces a team to change how it works has little chance of surviving once the project ends.
04 — What we learnedCode in service of the model
There is a way of doing educational technology that starts with the tool: “we have this platform, let’s see what we can put in it”. This project went the other way round. First the diagnostic, then the teaching model, and only then the code. The module exists because the model needed it, not because someone wanted to build something.
Three lessons I take with me:
- The data is the conversation. The diagnostic was not a box to tick. It was the moment the School, the teaching team and the technical team started speaking the same language.
- Personalising is not complicating. For the learner, the experience is simpler: log in, answer, and know what comes next. The complexity stays on the tool’s side.
- Building is also handing over. A tailor-made solution only makes sense if the institution can sustain it. That is why documentation and support weigh as much as the code.
05 — What changedEvidence over rank
Let’s go back to the two people from the start. With this model, the municipal judge and the person in the administrative area no longer get the same invitation. Each one completes her diagnostic and finds a path that starts from what she knows and points towards what her work demands of her.
Each person’s path is set by the evidence from their diagnostic, not by the org chart.
For the School, that means more than a good learning experience: it means knowing where the gaps are and being able to decide, with data, where to reinforce. For me, it is the best demonstration of something I believe in: technology is worth it when it gives shape to a good idea, not when it replaces one.
A project developed by Innovahub with the Rodrigo Lara Bonilla Judicial School of Colombia’s Superior Council of the Judiciary. The figures come from the diagnostic carried out between November and December 2025. The work was collective: the teaching team, the technical team and the School.