SUMMER INTERNSHIP AT DE SHAW

Reimagining Coresight for Rapid Incident Response

Reimagining Coresight for Rapid Incident Response

Role

Interaction designer (Solo)

Domain

Fintech | Enterprise Software

Duration

7 weeks

Non disclosure

This work showcase is a record of my internship project, focusing exclusively on the UX process and resulting design solutions. All data, metrics, and screenshots are blurred, generalized, filtered, or altered to protect proprietary information and honor the NDA.

About internship

During summer 2024, I worked as an interaction designer at D. E. Shaw & Co., a leading global hedge fund known for its expertise in quantitative trading.

My time at D. E. Shaw was highly collaborative and intellectually rigorous, focused on delivering a UX that would provide technical teams with maximum efficiency and system clarity.

Problem statement

"The Coresight application suffered from a cluttered, data-dense UI and broken information hierarchy, making complex workflows difficult for Ops/Dev stakeholders to identify and address system problems quickly. My project was to redesign and simplify the UX across the platform, including creating a new mobile version, to enable seamless, on-the-go analysis and rapid action."

What is CoreSight?

Monitor

Alert

Analyze

Act

Coresight is a real-time monitoring tool used by the firm's technical teams (Ops/Dev) to track the status of trading and support applications. Essentially, it's a dashboard that lets users see the health of critical systems as it happens.

PROBLEMS

Cluttered UI and Complex Workflows Slowed User Actions

Users couldn't quickly pinpoint critical problems.

Complex workflows required too many steps to take critical action.

The interface was overwhelmingly tough for new users.

Simple tasks were either impossible or too complex for users to figure out without assistance from support.

Off-hours monitoring wasn’t possible - needed a mobile version.

Screenshots from old UI

THE SOLUTION

Solving the Clutter Crisis: An Intuitive, Actionable UX

Status Alerts

*Designed for 27 inch monitor used by Devs/Ops

*Designed for 27 inch monitor used by Devs/Ops

Key Dashboard Improvements

Focused Clarity:

The entire view is designed for quick triage; product functions are separated into tabs, ensuring this screen focuses exclusively on essential resource alerts and features only simplified, critical navigation.

Intuitive Organization:

Alerts are intelligently grouped by dependency (Domain Groups) to instantly clarify the potential impact of an issue. The clean UI ensures that information hierarchy guides the user's eye to the most important data points.

Unified Action:

Action-taking workflows have been standardized and unified across all resources, solving complexity issues. The system fully supports efficient bulk actions on multiple selected resources simultaneously.

Key Dashboard Improvements


  • Clarity: Tabs divide the product, keeping this screen focused only on essential resource alerts.

  • Organization: Alerts are grouped by dependency (Domain Groups), clarifying the impact of issues.

  • Hierarchy: The UI is clean, featuring simplified navigation and only essential information.

  • Action: Action-taking is unified and standardized, supporting bulk actions on multiple resources simultaneously.

Resource Detail View

*Designed for 27 inch monitor used by Devs/Ops

*Designed for 27 inch monitor used by Devs/Ops

Key Detail View Features

Key Detail View Features


  • Organized Panel: A clean side panel template organizes all resource-specific information consistently, regardless of the resource type.

  • Problem Summary: A dedicated section clearly outlines what went wrong, providing the context needed for immediate investigation.

  • Unified Action: All possible actions (e.g., Acknowledge, Resolve, Add Observer) for the selected resource are available in one centralized location.

  • Essential Metrics: Users can view key performance metrics and charts (like CPU and Memory) instantly, supporting immediate, data-driven analysis.

  • Integrated Chatbot: A built-in chatbot allows users to ask quick questions about the resource, its history, or the alert status without leaving the page.

Organized, Unified Panel:
A clean side panel template organizes all resource-specific information consistently regardless of type. This panel includes a dedicated Problem Summary outlining what went wrong, providing immediate context for investigation.

Unified Action:

All possible actions (e.g., Logging ticket, Ignoring, Configuration, Copying Path) for the selected resource are unified in one centralized location, drastically reducing the effort spent navigating to take critical action.

Complete analysis:

Users can view essential performance metrics and charts (like CPU and Memory) instantly for data-driven analysis. Plus, an Integrated AI Chatbot allows for quick historical lookups or status checks without leaving the page.

All Resources Tab

*Designed for 27 inch monitor used by Devs/Ops

Key Dashboard Improvements

Key Dashboard Improvements


  • Dedicated View: A separate tab houses this view, ensuring the primary Alerts dashboard remains focused for daily monitoring (since this screen is rarely used).

  • Full Customization: The entire view is fully customizable by the user (size, position, and column visibility), addressing the need for efficiency by letting users see only what's essential for their role.

  • Type Organization: Resources are logically grouped by type (e.g., Processes, Hosts, Disks, etc.), allowing for quick, targeted searches and analysis.

  • Inline Visualization: Mini-graphs are embedded directly within the grid rows (e.g., CPU, Memory), enabling at-a-glance comparison and easy visualization of anomalies across many resources.

Customized Efficiency:
A dedicated, separate tab houses this view (as it's rarely used), maintaining focus on the primary alerts. The entire interface is fully customizable by the user (size, position, and column visibility), allowing teams to see only the essential data for their specific role.

Intuitive Organization:

Resources are logically grouped by type (e.g., Processes, Hosts, Disks), allowing for quick, targeted searches and analysis. This structure directly supports rapid investigation when moving beyond the initial alert.

At-a-Glance Insights:

Mini-graphs are embedded directly within the grid rows (e.g., CPU, Memory), enabling at-a-glance comparison and easy visualization of anomalies across many resources without needing to drill down.

Mobile Version

Strategy for Mobile Version

Strategy for Mobile Version


  • Alerts Focused: The design is a subset of the web application, prioritizing status and alerts for quick on-the-go analysis.

  • Scannable Alerts: Alerts are presented in large, simplified cards with key context (reason, severity) for rapid assessment.

  • Essential Actions: Only the most critical actions (Ignore/Acknowledge and Ticket Logging) are available, matching off-hours use cases.

Alerts Focused:

The design is a subset of the web application, prioritizing status and alerts for quick on-the-go analysis.

Scannable Alerts:

Alerts are presented in large, simplified cards with key context (reason, severity) for rapid assessment.

Essential Actions:

Only the most critical actions (Ignore/Acknowledge and Ticket Logging) are available, matching off-hours use cases.

Design process

1

Understanding

Secondary Research

  • Documentation review

  • Panopto Videos

  • Content Audit

  • Competitor analysis

Primary Research

  • Stakeholder Interviews

  • User Interviews

  • KT sessions

Secondary Research

  • Documentation review

  • Panopto Videos

  • Content Audit

  • Heuristic Analysis

  • Competitor analysis

Primary Research

  • Stakeholder Interviews

  • User Interviews

  • KT sessions




2

Analysis

  • Affinity Mapping

  • Personas

  • Journey Mapping

3

Design

  • Ideation

  • Info. Architecture

  • Wireframes

  • Final Screens

  • Prototyping

4

Testing and Iterations

Testing and Iterations

  • User testing

  • Stakeholder Feedback

  • Team Reviews

  • Design Iterations

Understanding

Auditing the system and User Workflows

Glimpse of content audit

Secondary Research

Focus:

  • User/Tool Context: Understood the Coresight's functionality and who the core users were and how their workflows look like.

  • System Usability: Conducted Heuristic Analysis to understand the existing application's usability position and current state.

  • Market Gaps: Analyzed competitor tools to determine where the product lacked in features and workflow efficiency.

Top Insights:

  • Confusing Architecture: The overall product structure was highly confusing, making it difficult to determine where key actions or processes were located.

  • High Barrier to Entry: The application was so cluttered and complex that the barrier for entry was excessively high for new or infrequent users.

  • Broken Hierarchy: Essential information was presented without any hierarchy, forcing users to learn multiple layers of unnecessary context just to understand basic functions.

Primary Research

5 user interviews

3 KT sessions with team

4 Stakeholder Interviews

Focus:

  • Workflow Mapping: Contextually and systematically map the users' current, complex workflows for alert handling and problem resolution.

  • Pain Points: Identify key usability barriers, specifically regarding information overload and the difficulty of taking critical action.

  • Mobile Needs: Define the essential subset of features and data required for a functional, on-the-go mobile monitoring experience.

Top Insights:

  • Inefficient Action: The existing system required an unacceptable amount of clicks and steps to perform core tasks, severely hindering efficiency.

  • Data Scarcity/Excess: Users wasted time either searching for critical data hidden within clutter or dealing with excessive, non-relevant information.

  • Wasted Time: Due to poor hierarchy and documentation, users frequently had to navigate multiple systems or create Jira issues to find answers, resulting in significant wasted time.

  • Mandatory Mobile: Off-hours monitoring was confirmed as a mandatory use case, proving the necessity of a dedicated, simplified mobile application for rapid assessment. Although, there was a constraint that the app must be on web, and not an app.

Design Phase

From Concept to Clarity: Ideation, Wireframes, and High-Fidelity Output

Ideation Sketches & Wireframes

Prototype

testing and iterations

Rapid Validation: Utilizing RITE to Converge on Final Solutions

I defined this phase using the Rapid Iterative Testing and Evaluation (RITE) methodology. To maximize efficiency, I conducted frequent, small-scale usability tests with users and had meetings with developers and stakeholders. Any critical usability flaws were immediately fixed in the high-fidelity prototype and re-tested as quickly as possible.

Key Iterative Decisions:

  • Strategic Feature Integration: I pivoted from building a new data analyzer to integrating existing internal tool for data visualization ("Mirror"). This decision saved significant development time and focused Coresight on its core strength: real-time alert handling.

  • Usability & Efficiency Refinement: I implemented multiple UI changes to speed up workflows, including revising the Problem Status Cards, optimizing filtering/searching, and simplifying the bulk action selection process.

  • Logical Problem Grouping: I redesigned the resource display to group problems based on their technical interrelation, correcting a hierarchy flaw. This grouping allowed me to integrate AI-driven insights based on that correlation, significantly speeding up the analysis.

Non disclosure

This work showcase is a record of my internship project, focusing exclusively on the UX process and resulting design solutions. All data, metrics, and screenshots are blurred, generalized, filtered, or altered to protect proprietary information and honor the NDA.

Thank You for scrolling by!