CloudGlance - Document Intelligence Platform

I shaped the structure that would later drive the entire product:

• Project = workspace model
• Files (global vs project-level)
• Tasks and Task Runs
• Indexing logic
• AI Assistant (global vs project-focused)
• Navigation hierarchy
• Output flows and Split-view system

Many ideas already existed directionally; my role was to clarify them, structure them, and turn them into consistent, usable patterns that would scale.

While designing early screens, it became clear that CloudGlance needed a cohesive identity, something stable, mature and enterprise-friendly.

This wasn't part of the original brief; I initiated it because the platform needed it.

Only after presenting the reasoning did the team agree that a new identity was essential.

The earlier logo leaned literal and playful, revolving around a bulky "cloud" visual and green palette. It didn't reflect the seriousness of the workflow nor the enterprise clients they were approaching. It also overlapped with their competitor landscape.

I suggested creating a new identity system that could:

• communicate maturity and trust
• visually hold the weight of enterprise workflows
• feel timeless, not startup-like
• live consistently across platform, website and future materials

The team welcomed the direction.

I designed a compact symbol built from C + G + a focal point - representing "glance", clarity and attention without being literal.

It reads cleanly at small sizes, aligns with enterprise expectations, and supports a more composed visual language.

I replaced the green palette with a composed, trustworthy blue.

Blue carries the exact qualities the platform needed:

• Stability
• Calm
• Reliability
• Professional maturity

Accents are used sparingly to maintain a calm environment for long work sessions.

Neutral, readable sans-serif.

Structured hierarchy.

Comfortable line spacing.

Nothing decorative, entirely built for clarity and long-form reading.

Calm over clever
No flashy AI theatrics. The design stays thoughtful, quiet and composed.

Predictable patterns
Same structures reappear everywhere - tabs, headers, panels, workflows.

Depth only when needed
A simple overview → deeper layers only when the user asks for them.

Designed for senior users
Clear labels, comfortable spacing, no hidden interactions.

Transparency over abstraction
Users always know what the AI saw, which files were referenced, and where answers came from.

Intentional empty states
Guiding but not decorated.

UI gets out of the way
Documents and outputs are the real focus.

Built for long work sessions
Soft neutrals, stable spacing, low visual fatigue.

Navigation that always shows "where you are"
One of the most important systems I designed:

• Sidebar has open/closed modes
• Logo doubles as the toggle to avoid cluttering headers
• When inside a project → a sub-item appears under "Projects"
• When inside a task output → a small drawer shows the task name
• In output-heavy screens, the sidebar auto-collapses to free horizontal space
• If reopened, it appears as a temporary overlay
• This gives both orientation and fast access without breaking the flow
• This spatial awareness system prevents disorientation, especially for older users and becomes one of the signatures of the platform's usability.

Designed to give users orientation without demanding attention.

The dashboard is intentionally simple. Users land here, orient themselves, and continue.
No analytics noise.
Just clarity, recent items and a stable sense of place.

Where teams organise, review and manage every tender in a structured way.

Two views (grid and list) serve two habits:

• Visual scanning
• Metadata-heavy review

Search, sort and filtering give teams control across dozens of ongoing tenders.

The layout mirrors familiar enterprise tools, reducing onboarding friction.

A clean snapshot that helps users understand the project before diving deeper.

A clean, high-level snapshot: recent activity, last runs, status and type.
No charts, no clutter - it's there to help users settle into the project before going deeper.

All documents live here - global or project-specific with clear indexing states.

Files are the foundation of CloudGlance's intelligence.

Two scopes, so users always know which scope they're in:

• Organisation Folder (global files)
• Project Folders (per project, narrowed)

Indexing states show whether the AI knows the document yet - building trust into the workflow.

Repeatable, reliable workflows built on top of Cortex’s backend intelligence.

Cortex had already built the task logic. I shaped the structure around it.

Key concepts:

• Task Catalog (what you can run)
• Task Runs (every execution saved for audit)
• Repeatable runs
• Processing states
• Clear steps → run → outputs → citations → rerun.

This structure supports all current and future automations without redesigning the product model.

An initiation flow designed to stay out of the way - fast, predictable, and always anchored to the user’s current workspace.

Certain actions in CloudGlance happen inside modals rather than full screens.
This keeps the platform focused, prevents unnecessary navigation jumps, and supports a predictable workflow rhythm.

Every automation begins with a consistent three-step structure:

• Select a Project
• Choose the task type
• Select input files

This creates a simple, repeatable pattern regardless of how complex the underlying workflow is.

But the modal adapts to where the user starts the action:

• From Dashboard: users see all three steps.
• Inside a Project: the first step is skipped; the modal starts at task type → files.

This reduces friction and shortens the path to action, especially for users who run tasks repeatedly within the same project.

The consistency of the modal keeps the platform scalable - new task types can be added without redesigning the initiation flow.

A dual-layer space built for quick validation and deep verification.

Outputs needed to support two behaviours:

• Quick validation
• Deep verification

So I designed a dual-layer view:

• Summary tab
• High-level answers for quick understanding.
• Split-view
• Left: extracted content
• Right: citations & source document

Users can verify any answer instantly.
During this mode, the sidebar auto-collapses to maximize screen space.

A conversational layer that pulls answers from exactly where the user needs.

Two modes:

• Global chat: searches across organisation-wide data
• Project chat: narrowed to the project (user can create a focused chat)

Chat history lives in the sidebar; Inside a project, users only see project-focused chats.
Answers are cited, trustworthy and direct.
The design stays calm, no "conversational UI gimmicks."

A dependable audit trail that keeps teams aligned and accountable.

A clean activity timeline that groups events and actions.

Important for audit trails, approvals and multi-team workflows.

Simple, trustworthy entry points for a platform handling sensitive data.

The peripheral screens in CloudGlance follow the same design philosophy as the main product - calm, unobtrusive and clarity-first.

Login & Signup
Minimal layouts, neutral surfaces and straightforward forms establish trust from the first interaction.
Nothing performs theatrics; the focus is simply on getting users into the platform with confidence.

Settings (Modal-Based)
Account settings and lighter configuration actions open in small modal panels rather than taking users away from their workspace.
This keeps the platform feeling continuous and prevents context loss during long work sessions.

Secondary screens stay intentionally quiet, their job is to support the workflow, not take attention away from it.