Tools and Technology

A breakdown of the platforms, APIs, development tools, and infrastructure powering this AI aggregation framework.

Dissertation Timeline

Frontend Tools

  • HTML5 & CSS3: Structured content and consistent layout styling.
  • JavaScript (vanilla): Dynamic behaviour across sections, collapsible elements, iframe loading.
  • Font Awesome: Semantic visual iconography for intuitive navigation.
  • Responsive Layouts: Optimised for both desktop and mobile viewing experiences.
  • Accessible Contrast: Ensures readability and screen-reader compatibility.
  • Section-Specific Styling: Colour-coded panels using class-based structure.

AI APIs and Frameworks

  • OpenAI (GPT-4): Participated in synconatic questioning across all scenarios.
  • Gemini: Queried in parallel to assess model variation and response structure.
  • Grok (xAI): Included for cross-architecture comparison with other LLMs.
  • SHAP: Applied to highlight influential input features behind outputs.
  • LIME: Used to generate surrogate interpretability models.
  • MMLU / HELM: Datasets for assessing factuality, reasoning, and coherence.

Deployment & Development Environment

Scripting & Automation

  • Prompt Automation: Scripts for batch-model query and response structuring.
  • File Management: Sorting AI outputs for easy mapping to HTML views.
  • LangChain: Framework consideration for response chaining and logic mapping.

Accessibility Considerations

Inclusive design focused on clarity, keyboard navigation, and simplified structure for accessibility support.