Contestio: How Three Event Projects Became One Platform

When you build the same thing three times for three different clients, you stop and ask yourself a question: why isn't this a product yet?

That question, asked in late 2024, is what produced Contestio — our end-to-end contest management platform for communities and organizations. But the story starts in early 2022.

PRiDA İletişim Ödülleri (February–April 2022)

The first project was a Turkish communications industry awards program. The requirements: an event website, a content management interface for the organizing team, a submission management system for applicants, and a jury evaluation module so judges could score entries online.

We built it from scratch in PHP and HTML. It worked. The award ceremony ran smoothly. We moved on.

PSM Awards (March–July 2022)

A few weeks later, a second client came with a similar brief. Another industry awards program. Same requirements: event site, content management, registrations, jury scoring, result publication.

We built it again. Different design, same structure. We recognized the overlap but treated it as coincidence. It wasn't.

Huawei Seeds for the Future Turkey (March–July 2024)

Two years later, we were contracted to build the digital infrastructure for Huawei's "Seeds for the Future" technology program in Turkey — a competition for university students. The requirements were, once again, the same: event website, registration management, content management, voting system.

By this point the pattern was undeniable. Three clients across three years had come to us with nearly identical needs. Each time, we had built a custom system. Each time, we had spent engineering time on the same foundational modules: user registration, application intake, jury scoring logic, result tallying, announcement publishing.

The Decision to Productize

After the Huawei project, we wrote down everything these three systems had in common. The list was long: application management, jury assignment and blind-review scoring, multi-stage evaluation pipelines, result publication with access controls, statistical reporting, email notifications at each stage.

The differences between the three projects were mostly visual — each client had its own branding, its own category structure, its own deadline schedule. The logic was the same.

Contestio was built by taking the Huawei system as the most recent and most mature version, extracting everything client-specific into a configuration layer, and building a multi-tenant architecture around it. An organization can now run an awards program or competition without commissioning custom software. They configure categories, set deadlines, invite jury members, and Contestio handles the rest.

Jury Assignment: The Algorithm Problem

One of the harder design decisions in Contestio was jury assignment. Manually assigning jury members to submissions is straightforward in small competitions — you have ten jurors and thirty entries, you hand-distribute them. At scale, and when you need to guarantee fairness properties, it becomes genuinely difficult.

Contestio ships two jury assignment algorithms that organizers can choose between depending on the nature of their competition.

The equal distribution algorithm ensures that each juror reviews approximately the same number of submissions. This is the right approach when you want workload fairness — no juror is overwhelmed while another reviews only a handful of entries. It's also the right approach when all jurors are considered equally qualified to evaluate all submissions.

The shared pool algorithm works differently: submissions are placed in a common queue and jurors draw from it. This approach is better when you want cross-validation — ensuring that each submission is seen by multiple independent reviewers without pre-assigning who sees what. It naturally handles situations where jurors have different availability windows, since each juror pulls from the queue at their own pace.

Both algorithms respect conflict-of-interest exclusions: if a juror is affiliated with an applicant organization, the system automatically excludes that juror from evaluating that submission.

Scoring: Three Models for Different Competition Types

Not all competitions score the same way. Contestio supports three scoring system types to cover the range of real-world needs.

School-based scoring maps jury evaluations to letter-grade equivalents — useful for educational competitions where outcome communication needs to align with academic grading conventions that participants already understand.

Grade average scoring computes a weighted average across evaluation criteria. Each criterion can carry different weight. The final score for a submission is the weighted mean of all juror evaluations across all criteria. This is the most common model for professional awards programs.

Project-based scoring treats each submission as a project with distinct deliverable components, each scored independently. The aggregate score can reflect different weights for different deliverable types — useful for competitions where some components (a written proposal, a prototype, a presentation) matter more than others.

Automatic result calculation runs as soon as the evaluation period closes. The system tallies scores, applies the configured weighting model, and generates ranked results without manual intervention.

AI Scoring Support

Contestio includes an optional AI scoring support tool. It's worth being precise about what this means — and what it doesn't mean.

The AI support tool is an analytical assistant for jurors, not an automated judge. When a juror is evaluating a submission, they can activate the AI analysis, which reads the submitted materials and produces a structured summary: key strengths, potential weaknesses, areas that align strongly with the competition criteria, and areas that may need closer examination.

The juror reads this analysis and makes their own scoring decision. The AI output is advisory — it's designed to help a juror quickly orient to a submission they're unfamiliar with, not to replace their judgment. The final score recorded is always the human juror's score.

This distinction matters. In high-stakes competitions — awards programs, scholarship selections, procurement evaluations — the legitimacy of the outcome depends on human accountability. AI can be a useful analytical tool in that process without becoming the decision-maker.

User Roles

Contestio supports multiple user roles that map to how competitions are actually run: competition administrators who configure and manage the overall program, category managers who oversee specific competition tracks, jurors who evaluate submissions within their assigned scope, applicants who submit entries and track their status, and public viewers who can see published results and announcements.

Role-based access controls ensure that jurors can't see each other's scores during the evaluation period, applicants can't see evaluation criteria weighting, and results remain private until the organizer chooses to publish them.

What Contestio Is For

Contestio is an end-to-end platform for communities and organizations running structured competitions: industry awards programs, scholarship and grant competitions, innovation challenges, university project competitions, and public sector procurement evaluations that involve a formal jury process.

It handles the full lifecycle — application intake, jury assignment, blind-review scoring, automatic result calculation, and result publication — with the configurability needed to match different competition structures. The AI scoring support tool is available for competitions where submission volume makes individual juror review intensive.

This is the thesis behind every Talivio product: real client work reveals real requirements. Contestio exists because we built the same system three times and finally had the sense to turn it into something reusable.