How to Write an AI Award Submission Judges Understand

November 15, 2026 · 6 min read

To write an AI award submission that judges understand, tell a clear story in four parts: the problem you set out to solve, the approach you chose and why, the outcomes you measured, and how you deployed the system responsibly. Use plain language, define any technical term you cannot avoid, and support every claim with evidence. A strong AI initiative can be undermined by an entry that hides its value behind jargon or makes claims it cannot back up. This guide walks through each part and the common mistakes to avoid.

Why plain language matters to an expert jury

At the Velto European Awards, nominations are evaluated by an independent Jury on the basis of professional merit and documented achievements. That phrase is worth keeping in mind as you draft. Jurors need to understand what was achieved and see that it is documented; they should not have to decode your internal vocabulary to get there.

Expertise does not mean every reader shares your specialism. Someone with deep knowledge of digital transformation may not know your model architecture, your sector's regulations or your company's acronyms. Writing in plain language is not simplifying your work; it is making its significance visible. If a sentence only makes sense to your own engineering team, rewrite it.

Before you start, read the AI & Digital Transformation Awards programme page and choose the nomination that best fits your initiative. Each programme comprises several nominations, and shortlists and winners are determined within each nomination, so a precise fit helps the Jury assess your entry against the right peers.

Start with the problem, not the technology

Many AI entries open with the model, the stack or the data pipeline. Judges usually need something else first: why did this matter? Open with the problem in terms anyone in your organisation would recognise.

  • Who was affected? Customers, clinicians, analysts, field teams, suppliers.
  • What was going wrong or missing? Slow decisions, inconsistent quality, manual work that did not scale, risks that were hard to spot.
  • What did it cost? Time, money, errors, missed opportunities, as you measured them before the project.
  • Why was AI a sensible answer? Explain briefly why a simpler fix was not enough.

A good test: could a capable colleague from finance or operations read your first paragraph and explain the problem back to you? If not, keep editing.

Explain your approach without the jargon

Once the problem is clear, describe what you built or changed. Focus on decisions and reasoning rather than a list of tools.

Describe what the system does, then how

Lead with function: "The system reads incoming service requests and routes them to the right team, flagging urgent cases for a person to review." Then, if relevant, add one or two sentences on the method. The Jury can follow the logic without needing every technical detail.

Define terms you cannot avoid

Some technical words are necessary. When you use one, define it in a short clause the first time. Avoid stacking acronyms, and do not assume that widely used industry terms mean the same thing to every reader.

Show the choices you made

Judges tend to value evidence of sound judgement. Mention the alternatives you considered, the trade-offs between accuracy, cost, speed and explainability, and how you involved the people who would use the system. This shows leadership and strategy, not just engineering.

Present measured outcomes, and show how you measured them

This is where many otherwise good entries lose strength. Outcomes should be specific, measured and attributable to your initiative.

  • Give a baseline. A result means little without a starting point. State what was measured before the change and over what period.
  • Explain the method. Say how the figure was calculated, over what timeframe and on what scope (one team, one country, the whole organisation).
  • Separate model metrics from business outcomes. Technical accuracy is useful context, but judges usually want to know what changed for people and the organisation as a result.
  • Be honest about attribution. If other changes happened at the same time, say so and explain how you isolated the effect of the AI initiative.
  • Include qualitative evidence where it helps. User feedback, adoption by teams or changes in working practice can support the numbers, but they should not replace them.

Where you can, attach or reference supporting documentation. Documented achievements carry more weight than assertions, and a short, well-labelled appendix is easier to assess than a long narrative full of figures.

Show responsible deployment

For AI initiatives, how you deployed the system is part of the achievement. A clear account of responsible practice signals maturity and reduces doubts a reader may otherwise have. Consider covering:

  • Data: where it came from, how you handled personal or sensitive information, and how you checked its quality.
  • Fairness and risk: how you tested for uneven performance across groups or situations, and what you changed as a result.
  • Human oversight: where people review, approve or override the system's output.
  • Transparency: how users and affected people are told that AI is involved, and how they can raise concerns.
  • Monitoring: how you track performance after launch and what happens when it drifts.

You do not need to claim perfection. Describing a limitation you found and how you addressed it is often more convincing than a statement that the system has no weaknesses.

Avoid unsupported claims and other common pitfalls

Confident language is fine; unsupported claims are not. Phrases such as "industry-leading", "revolutionary" or "first of its kind" invite scepticism unless you can prove them. Replace each superlative with the evidence that made you want to use it.

Other pitfalls to watch for:

  • Describing plans as results. Keep future roadmap items clearly separate from what has already been delivered.
  • Pilot figures presented as full-scale impact. State the scope of every result.
  • Marketing copy pasted into the form. The Jury is assessing merit, not promotional tone.
  • Too much detail in the wrong place. Put technical depth in supporting material and keep the main narrative readable.
  • A missing team story. Explain who led the work and how different functions collaborated.

Before submitting, ask someone outside the project to read the entry and highlight anything unclear or unproven. Then review the judging process and the published criteria so your structure reflects what the Jury will be looking for.

FAQ

When are the deadlines for the AI & Digital Transformation Awards?

For the 2027 cycle, the Early Bird Submission Deadline is June 13, 2027; the Regular Submission Deadline is July 14, 2027; the Late Submission Deadline is August 14, 2027; and the Final Submission Deadline is August 30, 2027. The Judging Period is August 31 – September 13, 2027, and Winners are Announced on September 14, 2027. Always check the Awards Timeline for current dates.

How much does it cost to enter?

Entry fees depend on the programme and the submission stage, and they increase with each stage. Current fees are listed on the apply page.

How many winners are there in the AI programme?

Each programme comprises several nominations. As a rule, there are three to five finalists and one winner (or none) per nomination, so the programme may have several winners across different nominations.

Ready to submit your entry?

If your AI initiative has a clear problem, a reasoned approach, measured outcomes and a responsible deployment story, it is worth putting in front of the Jury. Review the stages and fees and start your entry on the Apply now page. For questions, contact [email protected].