The Judge Experience
Great judging is its own kind of excellence. AwardScience makes it effortless.
Your judges are giving you something valuable — their expertise, their time, their reputation. AwardScience gives them a portal that respects all three: focused, fast, and designed around how expert evaluators actually work.
01 · Taking on the work
Your Personal Review Queue
When you log in, your dashboard shows exactly what’s assigned to you at your current stage. Nothing more, nothing less. Open any submission to read the full application in a clean, formatted view. Everything is laid out so you can focus on evaluating, not hunting for information or figuring out what you’re supposed to be looking at.
Accept, Decline or Declare a Conflict
Invitations arrive with a clear deadline, and you read the application before committing to anything. Accept to take it on, decline to pass it along, or declare a conflict of interest with an optional note for the administrator. Declaring a conflict frees your place for the next reviewer straight away, and every response is recorded against the round, so the panel’s composition stays defensible.
One Login Across Programs
Judge for several organisations with one email address and one password. Your credentials work across every AwardScience workspace you belong to, so there’s no separate account to remember for each program you sit on. Only you can change them — if an administrator needs to help, they send you a secure set-password link and never see or set your password themselves.
02 · Reading the entry
Blind Review, Built In
Personal details are removed before an application reaches you. Names, email addresses, phone numbers, links, ID numbers, places and organisations are scrubbed from what you see, and applicants can appear under an anonymous label instead of their own name. You judge the work, not the person — and afterwards the programme can show exactly which details reviewers could and couldn’t see.
Every Attachment Scanned
Open applicants’ documents with confidence. Every uploaded file is checked for malware before anyone can open it, and nothing can be downloaded until it passes. Files still being scanned are marked as such, and anything found to be unsafe is blocked outright rather than quietly left in place for someone to click.
Flag & Tag as You Go
Mark submissions with a red, yellow, or green flag as you work through your queue. Add your own tags to organize entries the way you think. Every note stays attached to your account with a timestamp, so your triage logic is yours alone and won’t get mixed up with anyone else’s. You can come back to flagged entries at any point during the review window.
03 · Scoring with AI help
AI-Assisted Scoring
Before you score, the platform proposes a score with written reasoning attached. You’re not handed a number to rubber-stamp. You’re given a first opinion to react to, with the logic visible so you know exactly what the AI weighted. Read the reasoning, enter your own score, and move on. Or disagree with it entirely and score from scratch. Either way, you submit the final number and that’s what counts.
Ask AI About Any Answer
On any field in the application, hit Ask AI to open a chat window. Summarise a long essay, ask a follow-up question, or pull out specific claims without leaving the review screen. It works inline, so there’s no tab-switching or copy-pasting. Useful when an entry is dense, and you need to cut through to what actually matters before scoring.
Catch Look-Alike Submissions
Similarity detection compares submissions by meaning, not just shared words. When two entries are making the same argument in different language, the platform surfaces them side by side with a similarity score. Keyword searches miss this kind of overlap. The semantic comparison catches it, and in high-volume programs with hundreds of entries, that’s not something you can reliably do by hand.
04 · Deciding as a panel
Comment on Any Answer or Document
Highlight a sentence in an answer, or drop a pin on a page of an uploaded document, and start a thread about it. @mention the people responsible for a stage, reply in line, and save a draft until you’re ready to post. Comments appear live for the rest of the panel, anchored to the exact words or the exact page they refer to, so nothing is lost to a side conversation.
Vote & Rank
In people’s choice or panel voting rounds, you cast votes directly from a gallery view. One tap per entry. There’s no separate process and no spreadsheet to manage. Your allotted votes are tracked automatically, so you always know how many you’ve used and how many remain.
Spot Disagreement Instantly
When multiple judges have scored the same submission, a judge anomaly view ranks those entries by how far apart the scores are, highest divergence first, measured by standard deviation. Panels can work through disagreements systematically rather than discovering them after the round closes. It makes those conversations easier to have, because you’re starting from the data rather than going by memory.
See the Whole Score Picture
Score summaries break down results by stage and by judge, with calculated totals. You can see not just where an entry landed, but how different reviewers got there, what each stage contributed, and how the final number was built. Useful when a score feels surprising and you want to understand what’s behind it before the panel discussion.
05 · Keeping on schedule
Track Your Progress
A progress view shows how many reviews you’ve completed and how many are still open. Straightforward, but genuinely useful when you’re working through a large queue and want a clear picture of what’s left before the scoring window closes. No digging through entries to count manually.
Reminders That Reach You
When reviews are waiting, reminders go out by email, WhatsApp, SMS or mobile push — whichever channel you actually check. Program teams choose exactly who is reminded, who is copied, and how often, repeating on a schedule until the stage is done. If a message can’t be delivered, AwardScience retries it and then switches channel, so a bounced email doesn’t quietly turn into a missed deadline.
Secure By Design
Your data is hosted on enterprise-grade infrastructure with daily backups and strict access controls.
Built on ISO 27001-Certified Infrastructure
AwardScience runs on AWS, which holds ISO/IEC 27001 and PCI DSS certifications for its infrastructure. Our own security programme is built around the ISO/IEC 27001 control domains.
GDPR-Ready Data Handling
EU data residency, Standard Contractual Clauses for international transfers, configurable retention and deletion, and a Data Processing Agreement available for enterprise customers.
Encrypted Data Storage
All data is encrypted at rest and in transit using AES-256 encryption. Hosted on enterprise-grade infrastructure with daily automated backups and strict role-based access controls.
See AwardScience through a judge’s eyes.
Your best evaluators will want to come back next year.
Choosing award judging software
If you are evaluating platforms rather than sitting on a panel, the judging stage is where the differences between them actually show. Three failures account for most of the trouble, and none is obvious while a programme is running.
Scope leakage. A judge can reach entries they were never assigned, or scores from a panel they do not sit on. Nobody notices until an entrant asks who reviewed their submission.
Undetected disagreement. Two judges score the same entry 9 and 4. Averaged, it becomes a middling 6.5 and disappears into the middle of the table. The disagreement was the signal, and the arithmetic erased it.
No decision record. Six months later someone asks why an entry was shortlisted. The scores exist, but the reasoning lived in a panel conversation nobody wrote down.
Five things to test before you commit
Any demo looks clean. These are worth checking yourself, in a trial, with your own entries — they are where platforms differ, and the differences only show at volume.
- Log in as a judge, not an administrator. Can you reach an entry you were not assigned by editing the URL? Administrators rarely check this, and it is the most common governance hole.
- Score the same entry from two accounts, far apart. Does the platform surface the divergence, or quietly average it away?
- If it offers AI scoring, ask it why. A score with no reasoning attached cannot be defended to an entrant, a board or a regulator.
- Submit two entries that argue the same thing in different words. Keyword matching will miss them. Semantic comparison will not.
- Export the decision record. Not the scores — the flags, the remarks, who left them and when. If that export does not exist, the record does not exist.
Where judges use the scale differently — one marks generously, another never gives above 7 — score normalisation is a separate question worth settling before you weight a panel. Our RFP guide applies the same discipline across the whole platform rather than just this stage, and the award judging process guide covers how panels are usually structured, independently of any software.
Video library
Watch it in action
See the judging workflow end to end — blind review and scoring, and how scores and results are visualised afterwards.
Explore by role
The same program, from every seat
One award program touches applicants, judges, program staff, sponsors and auditors. Each of them gets their own view of AwardScience — here are all six, with the one you are reading marked.