Clara Cheung / Work
01 / 05

Helix Atlas

Making complex software easier to understand and work with.

The problem

Teams need to understand large, unfamiliar codebases before changing them. The product has to turn complex analysis into something people can navigate, question, and share.

My contribution

As the primary designer, I shaped the UX and implemented full-stack workflows, from onboarding and graph exploration to progress recovery and document exports.

The outcome

Across my first year on Helix, I delivered 200+ feature, reliability, and UX improvements across 12 core workflows. The wider platform is used internally by 1,400+ colleagues.

My role

At 3Pillar, I designed and built Atlas workflows, then refined them based on user and client feedback. I also contributed to chat, document management, sharing, analytics, reliability, and recovery across Helix.

About Atlas at 3Pillar ↗

Where I focused

  1. Make complexity navigable

    Design graph and knowledge interfaces that help users follow relationships without losing their place or the larger context.

  2. Design for work in progress

    Handle long-running tasks, recovery, and restored progress so the interface reflects what the system is actually doing.

  3. Carry the result forward

    Support HTML, DOCX, and PDF exports, custom branding, and secure sharing so outputs fit into existing workflows.

  4. Refine in production

    Use feedback to revisit confusing interactions and critical workflows, then implement and validate the changes.

Selected results across Helix

86.7%

Smaller initial app shell bundle

Reduced the code needed for the initial application shell through targeted optimizations.

200+

Feature, reliability, and UX improvements

Across 12 core product workflows during my first year on the team.

1,400+

Internal users of the Helix platform

Company-wide product reach; these figures describe the wider Helix work, including Atlas.

Design engineering decisions

  • Long-running work needs more than a loading indicator. I worked on progress restoration, recovery logic, and the underlying data handling so people could understand and resume their work.

  • Owning both the interface and its implementation let me address friction across the whole workflow, including cases that only surfaced through production use.

Knowledge sharing and onboarding

  • I also created executive demos, board materials, and a company-wide training program covering 40 platform sections.

  • For one two-day turnaround, I delivered a presentation deck, character animation, and executive summary site, drawing on both my software and production-art experience.

Key takeaways

The experience is shaped by everything around the product’s capabilities and solutions: how people get started, understand its progress, recover from problems, and use the result.