Good Doctor – Workbench Enhancement

Our Process: Design Sprint

We used a sprint-inspired 6-phase process to break down, explore, and rebuild the Workbench experience.

Phase 1 – Understand

We interviewed internal doctors (dr. Adhiatma, dr. Elizabeth, dr. Ega) and analyzed logs from 34+ users. Key challenges:

• Patient data not visible during long scrolls

• Chat input and diagnosis fields unclear

• Doctors missed queues or patient messages due to poor alerts

• ICD X system didn’t support quick diagnosis input (only keyboard-based)

We organized the problem around three core challenges:

1. Efficiency in consultations

2. Scalability in queue and prescription handling

3. Reliability of drug and diagnosis entry

   
   

Phase 2 – Define

We mapped the full Doctor Journey across:

• Opening clinic

• Attending patient queue

• Viewing health files

• Entering complaints & diagnoses

• Sending prescriptions

Each step revealed UX friction and emotional stress:

• Frustration from diagnosis errors

• Overload during multitasking

• Uncertainty in medication availability

We then created 3 doctor personas and reframed the redesign with real pain points and expectations.

Phase 3 – Sketch

We divided our focus into 3 main areas:

1. Patient Queue & Info Panel

2. Consultation & Diagnosis Flow

3. Prescription Entry & Drug Lookup

Doctors co-sketch with us using whiteboards and live feedback. Ideas included:

• Sticky patient info headers

• Real-time “queue alert” badge

• Auto-fill ICDX with mouse click

• Better separation of input fields for allergy, diagnosis, and usage

  
  

Phase 4 – Decide

We voted based on technical feasibility and user impact, then storyboarded:

• New layout for sticky patient profile during scroll

• Smart auto-complete for ICD X input (clickable, not keyboard-only)

• Drug search that filters results based on active pharmacy inventory

• Inline diagnosis tagging that updates fast reply templates

  
  

Phase 5 – Prototype

We built a low-fidelity interactive prototype using Figma and Overflow, simulating:

• Full consultation from queue → diagnosis → prescription

• Sticky header, real-time notification, and simplified fields

• Inline prescription without pop-ups

  
  

Phase 6 – Validate

We tested with 5 internal doctors in real scenarios:

• Prescribing for fever, cold, and pregnancy cases

• Entering complaints and tags

• Managing queues while responding in chat

Feedback Summary:

• 100% said the new layout felt more structured

• 60% completed prescriptions in under 4 minutes

• All users preferred mouse + type input for ICDX over keyboard-only

• Drug stock visibility during prescription saved 2–3 minutes per session

UX Strategy & IA

We designed with 3 goals:

• Reduce errors by clarifying structure

• Support mental flow of clinicians under time pressure

• Make queue and drug data more visible upfront

UX Principles