From manual triage to
From manual triage to
AI-assisted prioritisation
AI-assisted prioritisation
Engineers in regulated environments lose hours triaging fragmented data across legacy systems. I designed a secure, multi-agent AI dashboard that translates complex queries into plain-language insights and dynamic visualisations, helping teams identify and resolve operational bottlenecks faster.
Engineers in regulated environments lose hours triaging fragmented data across legacy systems. I designed a secure, multi-agent AI dashboard that translates complex queries into plain-language insights and dynamic visualisations, helping teams identify and resolve operational bottlenecks faster.
Role
Lead UX/UI designer
Timeline
6–8 weeks
Team
Cross-functional team
Type
AI Product Design / Proof of Concept
Role
Lead UX/UI designer
Timeline
6–8 weeks
Team
Cross-functional team
Type
AI Product Design / Proof of Concept
Role
Lead UX/UI designer
Timeline
6–8 weeks
Team
Cross-functional team
Type
AI Product Design / Proof of Concept
Challenge
Engineering Queries (EQs) in a highly classified manufacturing environment were causing severe bottlenecks. With over a million documents scattered across legacy systems, engineers were losing hours to manual triage. Data dilution was causing schedule delays, threatening strict government SLAs, and driving significant staff stress.
Process
I joined after a six-month delay had caused project fatigue. I rebuilt alignment with stakeholders, clarified the AI assistant’s purpose, and helped the build lead define the scope before designing the UI. To design the right user experience, I worked closely with solution architects and AI engineers to understand the multi-agent configuration behind the product.
Solution
An AI dashboard powered by a task-specific, multi-agent architecture that lets teams query EQ and PI data in plain language. Built on a foundation of responsible AI, it features strict role-based access controls and dynamically generates charts to help leadership monitor bottlenecks.
Impact
The 4-week pilot proved the system could securely link EQs to source documents, categorise free-text, and generate shareable data visualisations. It demonstrated a credible, safe shift from manual triage to AI-assisted prioritisation, giving engineers faster answers and more time for high-judgement decisions.
Challenge
Engineering Queries (EQs) in a highly classified manufacturing environment were causing severe bottlenecks. With over a million documents scattered across legacy systems, engineers were losing hours to manual triage. Data dilution was causing schedule delays, threatening strict government SLAs, and driving significant staff stress.
Process
I joined after a six-month delay had caused project fatigue. I rebuilt alignment with stakeholders, clarified the AI assistant’s purpose, and helped the build lead define the scope before designing the UI. To design the right user experience, I worked closely with solution architects and AI engineers to understand the multi-agent configuration behind the product.
Solution
An AI dashboard powered by a task-specific, multi-agent architecture that lets teams query EQ and PI data in plain language. Built on a foundation of responsible AI, it features strict role-based access controls and dynamically generates charts to help leadership monitor bottlenecks.
Impact
The 4-week pilot proved the system could securely link EQs to source documents, categorise free-text, and generate shareable data visualisations. It demonstrated a credible, safe shift from manual triage to AI-assisted prioritisation, giving engineers faster answers and more time for high-judgement decisions.
Challenge
Engineering Queries (EQs) in a highly classified manufacturing environment were causing severe bottlenecks. With over a million documents scattered across legacy systems, engineers were losing hours to manual triage. Data dilution was causing schedule delays, threatening strict government SLAs, and driving significant staff stress.
Process
I joined after a six-month delay had caused project fatigue. I rebuilt alignment with stakeholders, clarified the AI assistant’s purpose, and helped the build lead define the scope before designing the UI. To design the right user experience, I worked closely with solution architects and AI engineers to understand the multi-agent configuration behind the product.
Solution
An AI dashboard powered by a task-specific, multi-agent architecture that lets teams query EQ and PI data in plain language. Built on a foundation of responsible AI, it features strict role-based access controls and dynamically generates charts to help leadership monitor bottlenecks.
Impact
The 4-week pilot proved the system could securely link EQs to source documents, categorise free-text, and generate shareable data visualisations. It demonstrated a credible, safe shift from manual triage to AI-assisted prioritisation, giving engineers faster answers and more time for high-judgement decisions.
