AI tool for customer support team

Contact

BrightSide A concept project exploring how AI can simplify and improve customer support tools.

My Role

  • Role: UI/UX Designer

  • Project Type: Personal Project (B2B SaaS Concept)

  • Duration: 1 Week

  • Platform: Web Dashboard

  • Tools Used: Figma, FigJam, Notion, ChatGPT (for prototyping AI logic)

Context & Problem

I noticed that many customer support platforms (like Salesforce Service Cloud) feel overwhelming — agents juggle multiple tabs, manually search through documents, and switch context constantly. These inefficiencies slow down response times and create frustration on both sides.
This project was a self-initiated exploration: I defined the problem, drew inspiration from existing tools, and applied Salesforce’s design system to imagine how AI could transform the experience.

Design Objectives

My goal was to reimagine a human-centered, AI-assisted ticketing system.



Key principles included:

  • Streamlined multi-ticket management

  • AI-powered information retrieval & suggestions

  • Minimal UI to reduce cognitive load

  • AI-driven task automation (reply suggestions, tagging, categorization)

  • Agent well-being at the core of the design

My Role

  • Role: UI/UX Designer

  • Project Type: Personal Project (B2B SaaS Concept)

  • Duration: 1 Week

  • Platform: Web Dashboard

  • Tools Used: Figma, FigJam, Notion, ChatGPT (for prototyping AI logic)

Context & Problem

I noticed that many customer support platforms (like Salesforce Service Cloud) feel overwhelming — agents juggle multiple tabs, manually search through documents, and switch context constantly. These inefficiencies slow down response times and create frustration on both sides.
This project was a self-initiated exploration: I defined the problem, drew inspiration from existing tools, and applied Salesforce’s design system to imagine how AI could transform the experience.

Design Objectives

My goal was to reimagine a human-centered, AI-assisted ticketing system.



Key principles included:

  • Streamlined multi-ticket management

  • AI-powered information retrieval & suggestions

  • Minimal UI to reduce cognitive load

  • AI-driven task automation (reply suggestions, tagging, categorization)

  • Agent well-being at the core of the design

Final Designs Screeens:

To demonstrate how the AI assistant (BrightSide) could integrate into existing support tools, I recreated simplified versions of a ticket list view and a dashboard view inspired by Salesforce. These served as the context for exploring agent workflows.

My Contribution

The main design focus was on the BrightSide AI tool — how it appears within these spaces, how agents interact with it, and how it reduces friction in daily ticket handling.

I designed the workflows for:

  1. Assistance in Case Management – Complex Issue Resolution

  2. Performance Insight & Goal-Setting Workflow

  3. Real-Time Sentiment Analysis – Proactive Customer Care

  1. Assistance in Case Management – Complex Issue Resolution

AI identifies similar past issue → BrightSide surfaces insights, including a relevant ticket and solution, and prompts the agent to draft a response.

AI drafts initial email → The assistant generates a ready-to-use response with product details and helpful links, editable by the agent.

Email added to ticket view → The drafted response is inserted into the chat for review, where the agent can still make changes. Attachments added before sending → Agent can search, attach, and review supporting documents before finalizing the reply.

Final review and send → The completed email with attachments is ready to send, ensuring faster and more accurate support.

  1. Performance Insight & Goal-Setting Workflow

The AI tool proactively analyzes performance data and prompts the user to set a target for the upcoming week.

The AI chat displays a clear overview of the user’s performance metrics, including the radar chart that shows how different KPIs have trended.

The user can now set a specific performance goal using an intuitive slider and input field before submitting their new target.

  1. AI-Powered Sentiment Analysis & Action

  1. AI-Powered Sentiment Analysis & Action

The AI tool flags a new message with negative sentiment, prompting the user for immediate action.

After the alert, the AI tool suggests two immediate action options—scheduling a call or drafting an email—to address the urgent issue.

To save time, the AI generates a personalized email draft for the user to review and send, reducing response time for urgent tickets.

As an alternative action, the user schedules a direct call with the customer, leveraging the AI to streamline the booking process.

Conclusion

To be done

© 2026 alberdbrahma.

© 2026 alberdbrahma.

Conclusion

To be done

© 2024 alberdbrahma.