Public Case Study

New York Times R&D · Eclipse

Keeping journalists reporting while professional video moves in the background

As UX/UI Designer, I designed the mobile workflows for selecting, annotating, and transferring large video files from the field. Proxy-first transfer, visible upload states, and recovery paths made a demanding process usable enough to test under reporting pressure.

Situation
Video journalists often had to stop filming and leave the scene to find Wi-Fi and a laptop before editors could receive professional footage.
My role
UX/UI Designer responsible for end-to-end interaction flows, interface design, prototyping, and design handoff to the mobile developer.
Contribution
  • UX design
  • UI design
  • Prototyping
Focus Area
Complex operational UX
Period
May 2021–January 2022
Verified change
Field testing informed the proxy-first transfer, parallel background upload, visible status, annotation, and recovery paths in the workflow.

Early evidence

Move useful footage first—and make every transfer state legible

This recreated workflow connects the core interaction to an approved prototype screen. It shows how journalists could select priority clips, add context, and continue filming while uploads ran, so the newsroom could receive useful proxy footage before larger source files.

01 · Ingest Connect the camera card Bring professional footage onto a dedicated Android phone.
02 · Reduce Create smaller proxy clips Send editorially useful footage before the full-resolution originals.
03 · Explain Select and annotate Attach names, captions, and context to the clips that matter.
04 · Transfer Upload in parallel and in the background Keep progress, failures, and recovery visible while reporting continues.
Newsroom receives workable footage sooner
Eclipse prototype screen showing four proxy files with encoding progress, estimated transfer time, and completed states
Prototype Approved Eclipse prototype frame from Sadi’s high-fidelity design work. No NYT field photography is reproduced.

Decision atlas

Three decisions made a difficult transfer workflow usable in the field

The design had to move useful footage sooner, make transfer activity clear under unreliable conditions, and keep a remote team aligned while requirements changed.

  1. 01 Send the useful footage before the largest file Full resolution → proxy → editorial access
  2. 02 Treat reliability as a visible interaction state Progress → interruption → resume
  3. 03 Use one prototype to freeze and hand off the workflow Alternatives → agreement → implementation

Consequential decision

Send editorially useful footage before the largest source file

The problem

Professional camera files were too large for dependable mobile transfer. Early field work showed that moving full-resolution source files could be slow and unreliable under reporting conditions.

Evidence and reasoning

Professional cameras could generate compressed proxy files alongside full-resolution footage. The smaller files retained enough quality for rapid editorial review while reducing the amount of data that had to move first.

Design response

I designed a proxy-first flow for importing footage, selecting and reviewing clips, and sending the material the newsroom needed first. The wider team chose Android and a dedicated secondary phone to support wired ingest and sustained transfer.

Proof

The documented workflow used proxy clips for early editorial access while retaining source files for later high-resolution use.

Source footageMaximum quality · very large filesPreserved for later high-resolution use.
Proxy firstSmaller file · useful nowPrioritized for mobile transfer and newsroom access.
Recreated transfer decision based on the documented Eclipse workflow
Read the NYT R&D workflow report ↗ Recreated

Consequential decision

Treat reliability as a sequence of visible, recoverable states

The problem

Background transfer allowed reporting to continue, but it could also hide activity. Field testing showed the need for clearer information about progress, interruptions, and what would happen next.

Evidence and reasoning

Field use exposed more than a speed problem: journalists needed confidence that several uploads were progressing, clarity when one failed, and a safe way to continue after an interruption.

Design response

I designed parallel, background upload flows with progress, estimated time, completion, error, and resumed-transfer states. Annotation and batch captioning kept editorial context with the clips rather than splitting it across calls and messages.

Proof

The programme iterated on parallel uploads, resuming partially completed transfers, and clearer status feedback. Those were wider programme changes informed by field testing, not implementation work attributable to Sadi alone.

QueuedClip and context ready
UploadingProgress and ETA visible
InterruptedFailure is explicit
ResumedContinue partial transfer
CompleteReady for newsroom use
Recreated state model; engineering implementation belonged to the mobile developer and NYT R&D team
Read the NYT R&D field-test report ↗ Programme-level learning

Consequential decision

Turn a changing remote brief into one implementable workflow

The problem

The team worked fully remotely between the US East Coast and Helsinki, with limited access to test users and requirements that changed through roughly four or five design iterations.

Evidence and reasoning

Text descriptions left too much room for competing interpretations. Concrete prototype alternatives made the full workflow, edge cases, and technical implications reviewable by stakeholders and the developer at the same time.

Design response

I maintained one high-fidelity Figma prototype as the shared source of truth. After reviewing changing requests through concrete flows, I proposed a constructive requirements freeze and refocused the journey on its essential transfer states.

Proof

The prototype connected weekly reviews, interaction decisions, and developer handoff in one artifact. It established the design boundary; it does not claim that I implemented the Android app.

Visual alternativesMake competing interpretations concrete
Weekly reviewTest the end-to-end workflow
Requirements freezeProtect an implementable core
HandoffOne source of truth for the mobile developer
Recreated collaboration model based on Sadi’s approved project record
Prototype

Results

What changed

Field testing gave the wider programme evidence to refine proxy-first transfer and the workflow’s reliability states. It tested the direction I designed, while engineering, field operations, testing, newsroom use, and programme outcomes remained the work of the wider NYT R&D and newsroom teams.

Evidence boundary

What this does not establish

This Case Study shows a field-tested design direction, not confirmed production deployment, programme impact attributable to Sadi alone, or ownership of the Android implementation. The app screen is an approved prototype artifact; recreated diagrams are explanatory, and NYT field photography is not reproduced.

Collaboration and constraints

Detailed interaction design inside a live R&D programme

My responsibility

I designed the end-to-end mobile flows and UI, explored alternatives, maintained the high-fidelity prototype, facilitated design reviews, controlled the interaction scope, and handed the agreed states to the developer.

Wider team

I worked with one mobile developer, NYT R&D stakeholders, and journalist/test users. Engineering, field operations, testing, newsroom use, and programme outcomes belonged to the wider team.

Material constraints

The Android-only MVP had to handle professional camera formats, large files, variable 4G/5G connections, background activity, heat and battery limits, interruptions, limited tester access, and an eight-hour time-zone difference.

Start a conversation

What has become fragmented, difficult to use, or hard to decide?

Tell me briefly about the situation. I’ll reply with an initial perspective and a sensible next step.

Email Sadi