About
Service Assistant is a fully autonomous AI Agent that replaces the traditional contact form. Solves 98% of customer questions automatically, easily integrates to customer's website and support tools.
Service Assistant is a fully autonomous AI Agent that replaces the traditional contact form. Solves 98% of customer questions automatically, easily integrates to customer's website and support tools.
Challenge
Solar technicians and homeowners often struggle to find precise answers about their equipment. From troubleshooting inverter error codes to locating installation diagrams. Traditional support channels force users to dig through dense PDF manuals or wait days for email responses, leading to frustrated customers and overloaded support teams.
Solution
An AI Service Assistant trained to instantly answer technical questions by analyzing solar companies’ proprietary documentation, including product manuals, spec sheets, and support websites.
Years
2024 – 2025
Scope of work
Wireframes
Prototyping
Product Design
Branding
Animations
Tools
Figma
Jitter
Customizable AI Agents
To ensure client company can deploy the AI Service Assistant in a way that best fits their workflow, I designed multiple integration formats. Each tailored to different use cases while maintaining a cohesive user experience.
Customizable UI integrates seamlessly into existing company websites, blending with brand aesthetics while delivering AI-driven support.
Flexible Design
By building a flexible design system I ensured the Service Assistant can be viewed across any interface, whether as a desktop app or an embedded chatbot.
Shared components like conversation bubbles, search bars, and document viewers maintain consistency, while platform-specific layouts optimize usability.
Wireframes
To refine the AI Service Assistant’s usability, I developed medium-fidelity wireframes balancing enough detail to test functionality while staying flexible for iteration.
I mapped critical paths like asking assistant a technical question, presenting external sources, following up with related questions or collecting feedback.
Interactive Prototypes
Figma prototypes across whole app and admin dashboard allowing user testing and efficient iteration.
Dashboard
Admin Panel allows clients to quickly and easily set up and configure their own Service Assistant AI instances and track their performance.
To align design and development, I worked closely with developers using Material UI as our shared framework. This ensured visual consistency while keeping implementation efficient. The result? Faster builds and a unified system that both teams could iterate on seamlessly.
PDF Manuals and Websites as a Response Sources
If a customer needs further information about their issue, Service Assistant AI can direct to specific pages in the producers manual where the answer is located or redirect user to a related website. Administrators can easily upload files and attach them to a specific device for an AI agent to learn from.
Modular Components
Using components based design system ensures the app is flexible, scalable and easily customizable by the client.
Light and Dark mode
Using variables allows to easlily utilize dynamic themes to validate accessibility, contrast, and visual hierarchy.
