Designing an AI platform for research proposal evaluation
Helping researchers and evaluators review whitepapers, understand AI-generated scores and decide what progresses.
Making AI-assisted evaluation usable
During a two-month freelance contract in 2023, I designed the UX/UI for a new extension to the Ideateplus platform. The tool helped users upload and manage whitepapers, review AI-generated due-diligence results and decide whether a paper should be submitted or taken forward.
My Role
My responsibility covered the Product Design across this side of the product: accounts, the dashboard, whitepaper upload, AI interaction behaviour, scoring screens and the review and submission journey. I designed how users encountered and worked with the AI within the platform, including the surrounding screens and actions needed to complete their task. The extension was designed to connect with the wider Ideateplus ecosystem, bringing the evaluation workflow into a platform experience that users could return to as their work progressed.
Making AI-assisted evaluation usable
During a two-month freelance contract in 2023, I designed the UX/UI for a new extension to the Ideateplus platform. The tool helped users upload and manage whitepapers, review AI-generated due-diligence results and decide whether a paper should be submitted or taken forward.
My Role
My responsibility covered the Product Design across this side of the product: accounts, the dashboard, whitepaper upload, AI interaction behaviour, scoring screens and the review and submission journey. I designed how users encountered and worked with the AI within the platform, including the surrounding screens and actions needed to complete their task. The extension was designed to connect with the wider Ideateplus ecosystem, bringing the evaluation workflow into a platform experience that users could return to as their work progressed.
Making a demanding review process easier to navigate
Growing submission volumes were putting pressure on a manual whitepaper review process. Reviewers needed to assess detailed proposals against relevant subjects and challenges, while researchers needed feedback they could act on.
In 2023, the proposed workflow required several AI models working together to produce the analysis and scores. My design challenge was to bring these outputs into one clear experience, helping users manage submissions, inspect the assessment and retain control over what progressed.
Making a demanding review process easier to navigate
Growing submission volumes were putting pressure on a manual whitepaper review process. Reviewers needed to assess detailed proposals against relevant subjects and challenges, while researchers needed feedback they could act on.
In 2023, the proposed workflow required several AI models working together to produce the analysis and scores. My design challenge was to bring these outputs into one clear experience, helping users manage submissions, inspect the assessment and retain control over what progressed.
Defining a clear relationship between AI and its users
I conducted user interviews to understand the existing evaluation process. Alongside comparative research, I explored three questions that guided the design.
How could reviewers understand the reasoning behind an AI-generated score?
Where would people need to review, edit or approve the AI’s output?
How could researchers use the feedback to improve their paper before submission?
I explored these questions through the dashboard, scoring breakdowns and summary editor, then played back the prototype with users to assess the proposed workflow and inform the next steps.
Defining a clear relationship between AI and its users
I conducted user interviews to understand the existing evaluation process. Alongside comparative research, I explored three questions that guided the design.
How could reviewers understand the reasoning behind an AI-generated score?
Where would people need to review, edit or approve the AI’s output?
How could researchers use the feedback to improve their paper before submission?
I explored these questions through the dashboard, scoring breakdowns and summary editor, then played back the prototype with users to assess the proposed workflow and inform the next steps.
Finding useful patterns beyond direct competitors
I reviewed adjacent products because the research did not identify a direct equivalent for the intended whitepaper workflow. Résumé assessment tools offered references for scoring and feedback, while document-processing platforms provided examples of submission management and review. I used these comparisons to define principles for the new evaluation experience.
Show the overall assessment alongside a breakdown, helping reviewers understand the findings behind the score.
Provide feedback on individual sections so researchers can identify improvements before submitting their paper.
Show evaluation status, pending reviews and required actions so users understand what is happening and what comes next.
Support human review of AI recommendations, with clear responsibilities for approval and feedback when a paper does not progress.
Finding useful patterns beyond direct competitors
I reviewed adjacent products because the research did not identify a direct equivalent for the intended whitepaper workflow. Résumé assessment tools offered references for scoring and feedback, while document-processing platforms provided examples of submission management and review. I used these comparisons to define principles for the new evaluation experience.
Show the overall assessment alongside a breakdown, helping reviewers understand the findings behind the score.
Provide feedback on individual sections so researchers can identify improvements before submitting their paper.
Show evaluation status, pending reviews and required actions so users understand what is happening and what comes next.
Support human review of AI recommendations, with clear responsibilities for approval and feedback when a paper does not progress.
From whitepaper upload to an informed decision
I connected document upload, evaluation tracking and detailed review within one workflow. Users could return to existing submissions, inspect the AI-generated assessment and refine the summary before requesting approval.
From whitepaper upload to an informed decision
I connected document upload, evaluation tracking and detailed review within one workflow. Users could return to existing submissions, inspect the AI-generated assessment and refine the summary before requesting approval.
What the work contributed
Across the two-month engagement, I designed the evaluation extension from document management through to scoring, detailed review and summary editing, supported by user interviews and prototype playback.
The project reinforced that presenting AI-generated analysis also requires designing how people question, refine and act on it. The interface needed to make the assessment understandable while keeping responsibility for the next decision with the reviewer.
What the work contributed
Across the two-month engagement, I designed the evaluation extension from document management through to scoring, detailed review and summary editing, supported by user interviews and prototype playback.
The project reinforced that presenting AI-generated analysis also requires designing how people question, refine and act on it. The interface needed to make the assessment understandable while keeping responsibility for the next decision with the reviewer.
Have an interesting challenge?
Let’s makesomething great.
Have an interesting challenge?