Platform
One system for the whole weekend, and the year around it
Not a form builder with a leaderboard bolted on. Every part knows which phase the event is in and what the others are doing — which is why windows close on their own and actions that no longer make sense are refused rather than quietly accepted.
01Registration and teams
From a join link to a full roster
People arrive through a link or an invite code. You approve them one at a time or switch on auto-approve and let everyone straight in. Teams form by code or by invite, inside the size limits you set, and formation closes on its own when the event starts.
Approvals with a queue
Pending, approved and rejected are states, not a spreadsheet colour.
Join codes and invites
A team shares its code, or the lead invites people by name.
Size limits enforced
Minimum and maximum team size are refused at the write, not flagged later.
Optional e-mail verification
Confirm addresses before anyone joins a team or submits.
02Running the event
The weekend, while it is happening
Announcements reach everyone at once with a severity, so an urgent notice does not read like a reminder. Teams raise help tickets carrying their table number and a phone number; mentors claim them, resolve them, or escalate to the organisers.
One message, everyone
Info, warning and urgent are distinct, and they look it.
Help queue that can be worked
Claim, resolve, escalate — nobody sits with a hand up.
Live activity
Which teams are active, which have gone quiet, who is online.
Volunteer tasks
Assigned lists, so the people helping know what to do.
Table 12 · phone on the ticket so a mentor can find them
Table 4 · phone on the ticket so a mentor can find them
Table 9 · phone on the ticket so a mentor can find them
03Judging and results
Scoring that survives being questioned
You define the criteria, their maximum score and their weight. Judges score what they are assigned; coverage is visible as it fills in. The board is the part someone will argue with, so it is the part that has to be right.
Weighted criteria
Nothing hardcoded — the rubric is yours.
Per-judge normalisation
A share of what that judge could have given, not a raw sum.
Visible coverage
How many judges scored each project, before it is announced.
No silent zeros
If scores fail to load, the ranking says so rather than showing everybody on nought.
Each judge is normalised against what they actually scored — grading half the criteria no longer drags a team down.
04Analytics and reporting
What happened, in a form you can send upward
Registrations over time, team sizes, submission rates, judging progress and drop-off between stages. Exports come out as CSV wherever there is a table, and past events stay browsable instead of being cleared out.
Programme analytics
Across every event, not one at a time.
Accreditation reporting
Per-criterion scores formatted for a faculty submission.
Archive that stays
Last year's event is still there, still readable.
Export everything
Participants, teams, submissions, scores — CSV, any time.
- Registered152
- In a team12884%
- Submitted9675%
- Fully judged8892%
05Access
Five roles, separate doors
Each role gets its own interface and its own permissions, enforced on the server rather than hidden in the UI. A judge cannot open the organiser dashboard, a mentor sees only the teams in their event, a participant sees only their own team.
Organiser
Runs the event end to end and sees everything.
Judge
Assigned projects and the scoring sheet.
Mentor
The help queue and the teams they support.
Participant
Their team, their submission, announcements.
Volunteer
The task list assigned to them.
See it with data in it
The walkthrough is the console populated with sample records: pipeline, accounts, people, events, analytics and reports, with the search and filters working.