PRODUCT DESIGN · 2022
AI Scribing System
AI Scribing System
AI Scribing System
What is this about
What is this about
Helping clinicians reduce documentation work without disrupting the consultation.
Helping clinicians reduce documentation work without disrupting the consultation.
Services
Product Design
Product Design
Employer
PIT Solutions
PIT Solutions
Platform
Web
Web
Type
HealthTech SaaS
HealthTech SaaS
This case study is based on design challenges I worked on professionally. Due to a confidentiality agreement, the client's branding, and UI have been redesigned, and certain details have been modified while preserving the design thinking and process.
This case study is based on design challenges I worked on professionally. Due to a confidentiality agreement, the client's branding, and UI have been redesigned, and certain details have been modified while preserving the design thinking and process.
Overview
Overview
Designed an AI-powered clinical scribing SaaS platform for doctors and clinics that helps automatically document patient consultations, organize medical records, and surface AI-powered insights from conversations and symptoms. The platform brings transcription, clinical note generation, patient documents, and AI assistance into a unified workspace, reducing the repetitive administrative work involved in documentation. The goal was to simplify complex clinical workflows while keeping doctors in control of reviewing, editing, and finalizing patient records, allowing them to spend more time focused on patient care.
Designed an AI-powered clinical scribing SaaS platform for doctors and clinics that helps automatically document patient consultations, organize medical records, and surface AI-powered insights from conversations and symptoms. The platform brings transcription, clinical note generation, patient documents, and AI assistance into a unified workspace, reducing the repetitive administrative work involved in documentation. The goal was to simplify complex clinical workflows while keeping doctors in control of reviewing, editing, and finalizing patient records, allowing them to spend more time focused on patient care.
Design Process
The Challenge
The Challenge
Clinical consultations require doctors to listen to patients, understand symptoms, make decisions, and document the encounter, often at the same time. Manually recording patient history, symptoms, diagnoses, and treatment plans adds significant administrative work to an already demanding workflow. The opportunity was to use AI to automate the repetitive parts of clinical documentation while keeping doctors in control of reviewing and finalizing the patient record.
Clinical consultations require doctors to listen to patients, understand symptoms, make decisions, and document the encounter, often at the same time. Manually recording patient history, symptoms, diagnoses, and treatment plans adds significant administrative work to an already demanding workflow. The opportunity was to use AI to automate the repetitive parts of clinical documentation while keeping doctors in control of reviewing and finalizing the patient record.
The question is
The question is
How might we reduce the burden of clinical documentation without making AI another thing clinicians have to manage?
How might we reduce the burden of clinical documentation without making AI another thing clinicians have to manage?
Understanding the users
Understanding the users
Before thinking about the interface, I first looked at who would use the product and what their responsibilities were. The platform was designed primarily for doctors and clinics, with different roles interacting with patient information throughout the care process. Doctors need to focus on the patient while documenting symptoms, diagnoses, and treatment plans, while clinical staff support patient records and documentation. Understanding these different needs helped define the workflows and information each role needed access to.
Before thinking about the interface, I first looked at who would use the product and what their responsibilities were. The platform was designed primarily for doctors and clinics, with different roles interacting with patient information throughout the care process. Doctors need to focus on the patient while documenting symptoms, diagnoses, and treatment plans, while clinical staff support patient records and documentation. Understanding these different needs helped define the workflows and information each role needed access to.
The Workflow
The Workflow
I mapped the different stages of a typical clinical consultation to identify where the biggest friction occurred. The workflow moves from patient consultation → conversation → transcription → clinical documentation → review and correction → patient record.
The biggest opportunity appeared between the conversation and the final documentation, where doctors have to stay engaged with the patient while simultaneously capturing important clinical information.
I mapped the different stages of a typical clinical consultation to identify where the biggest friction occurred. The workflow moves from patient consultation → conversation → transcription → clinical documentation → review and correction → patient record.
The biggest opportunity appeared between the conversation and the final documentation, where doctors have to stay engaged with the patient while simultaneously capturing important clinical information.
The landscape
The landscape
I looked at AI medical scribing products such as DeepScribe and Abridge to understand how they approach transcription and clinical documentation, while also studying patient-management platforms such as OpenEMR to understand workflows around patient registration, records, appointments, and clinical information. Rather than simply replicating existing patterns, I focused on understanding how these workflows connect, where they overlap, and where fragmentation creates unnecessary complexity for doctors and clinical staff.
I looked at AI medical scribing products such as DeepScribe and Abridge to understand how they approach transcription and clinical documentation, while also studying patient-management platforms such as OpenEMR to understand workflows around patient registration, records, appointments, and clinical information. Rather than simply replicating existing patterns, I focused on understanding how these workflows connect, where they overlap, and where fragmentation creates unnecessary complexity for doctors and clinical staff.
Insights
Insights
These were the key findings that shaped the direction of the product.
These were the key findings that shaped the direction of the product.
Documentation goes beyond transcription
Documentation goes beyond transcription
AI can capture a consultation, but doctors still need to turn that information into structured clinical notes.
AI can capture a consultation, but doctors still need to turn that information into structured clinical notes.
AI can capture a consultation, but doctors still need to turn that information into structured clinical notes.
Patient context matters
Patient context matters
Doctors need quick access to patient history, medical documents, reports, and prescriptions during a consultation.
Doctors need quick access to patient history, medical documents, reports, and prescriptions during a consultation.
Healthcare compliance
Healthcare compliance
Designing workflows that followed medical documentation standards while remaining easy to use.
Designing workflows that followed medical documentation standards while remaining easy to use.
AI needs to support, not replace
AI needs to support, not replace
AI-generated insights are useful, but clinicians need to review and validate information before it becomes part of the patient record.
AI-generated insights are useful, but clinicians need to review and validate information before it becomes part of the patient record.
Workflows need to stay connected
Workflows need to stay connected
Moving between transcription, documentation, patient records, and AI tools can create unnecessary context switching.
Moving between transcription, documentation, patient records, and AI tools can create unnecessary context switching.
Solutions
Solutions
Before diving into the visual design, I spent time wireframing the product to figure out what actually mattered to users. Instead of focusing on colors or typography, I concentrated on the flow, how a clinician would start a recording, read transcripts, review AI insights, and navigate between patients with as little friction as possible. Working with simple, low-fidelity wireframes made it easy to experiment, gather feedback, and refine ideas without getting attached to the visuals. It helped uncover usability issues early and gave everyone a shared understanding of the product before moving into high-fidelity designs.
Before diving into the visual design, I spent time wireframing the product to figure out what actually mattered to users. Instead of focusing on colors or typography, I concentrated on the flow, how a clinician would start a recording, read transcripts, review AI insights, and navigate between patients with as little friction as possible. Working with simple, low-fidelity wireframes made it easy to experiment, gather feedback, and refine ideas without getting attached to the visuals. It helped uncover usability issues early and gave everyone a shared understanding of the product before moving into high-fidelity designs.
Visual Designs
Visual Designs
With the core workflows defined, I moved into visual design with a focus on keeping the experience simple and easy to follow for clinicians. The consultation and live transcription remained the main focus, while important tools stayed close by. I designed SOAP Notes to turn the conversation into structured documentation, an AI Assistant to help with follow-up questions and clinical insights, and Patient Documents to keep reports and medical records accessible during the consultation. Bringing these together in one workspace reduced unnecessary navigation and made the overall workflow more efficient.
With the core workflows defined, I moved into visual design with a focus on keeping the experience simple and easy to follow for clinicians. The consultation and live transcription remained the main focus, while important tools stayed close by. I designed SOAP Notes to turn the conversation into structured documentation, an AI Assistant to help with follow-up questions and clinical insights, and Patient Documents to keep reports and medical records accessible during the consultation. Bringing these together in one workspace reduced unnecessary navigation and made the overall workflow more efficient.
AI-Generated Clinical Notes
AI-Generated Clinical Notes
To simplify documentation, designed a feature that automatically generates structured clinical notes from the consultation transcript. Clinicians can instantly create SOAP Notes or H&P (History & Physical) Notes, review the content, make edits, and export the final version. This reduces manual documentation time while ensuring notes remain organized, consistent, and easy to review.
To simplify documentation, designed a feature that automatically generates structured clinical notes from the consultation transcript. Clinicians can instantly create SOAP Notes or H&P (History & Physical) Notes, review the content, make edits, and export the final version. This reduces manual documentation time while ensuring notes remain organized, consistent, and easy to review.
AI Clinical Assistant
AI Clinical Assistant
To make the experience more helpful, we implemented an AI Chat assistant that lets clinicians interact naturally with patient conversations. Instead of manually searching through transcripts, they can ask follow-up questions, analyze symptoms, generate summaries, or clarify clinical information in seconds. By keeping everything within the same workspace, the AI helps reduce documentation effort and supports faster, more confident decision-making.
To make the experience more helpful, we implemented an AI Chat assistant that lets clinicians interact naturally with patient conversations. Instead of manually searching through transcripts, they can ask follow-up questions, analyze symptoms, generate summaries, or clarify clinical information in seconds. By keeping everything within the same workspace, the AI helps reduce documentation effort and supports faster, more confident decision-making.
Handover
Handover
Once the designs were finalized, I collaborated closely with developers to ensure the experience was implemented as intended. Detailed design specifications, reusable components, and interaction guidelines were shared through a structured handoff, making development more efficient and reducing ambiguity. Throughout implementation, feedback was incorporated, edge cases were refined, and the product was continuously reviewed to ensure the final experience stayed true to the original design.
Once the designs were finalized, I collaborated closely with developers to ensure the experience was implemented as intended. Detailed design specifications, reusable components, and interaction guidelines were shared through a structured handoff, making development more efficient and reducing ambiguity. Throughout implementation, feedback was incorporated, edge cases were refined, and the product was continuously reviewed to ensure the final experience stayed true to the original design.
Reflections
Reflections
Working on a medical AI platform was a new and rewarding experience. It gave me the opportunity to design for a domain where clarity, accuracy, and usability directly impact clinical workflows. Being part of a highly collaborative project with designers, developers, and stakeholders strengthened my ability to solve complex problems while delivering an experience that supports healthcare professionals in their daily work.
Working on a medical AI platform was a new and rewarding experience. It gave me the opportunity to design for a domain where clarity, accuracy, and usability directly impact clinical workflows. Being part of a highly collaborative project with designers, developers, and stakeholders strengthened my ability to solve complex problems while delivering an experience that supports healthcare professionals in their daily work.