Background

In 2024, I designed Altera AL's first consumer AI agent dashboard, making complex AI creation accessible to everyone. Through intuitive design patterns and simplified workflows, non-technical users could create and deploy their own AI agents. The project launched in 1 month with user-created agents reaching 100K+ minutes of gameplay, setting new standards for AI agents in gaming.

The site can be found here.

Problem

Altera AL needed a fresh approach to AI agent creation. Rather than presenting users with overwhelming forms and technical parameters, I focused on designing an experience that felt more like assembling building blocks. The challenge was making complex AI concepts feel natural and engaging, while ensuring users could create sophisticated agents without feeling overwhelmed by technical details.

Agent Creation Page 1

To make AI agent creation feel natural, I designed an onboarding experience that mirrors familiar chat interfaces and game quest systems. Users are guided through creation steps by an AI assistant, with progress tracked through quest-like objectives at the top. This approach transforms technical configuration into an engaging conversation, similar to popular AI chat tools but with clear game-inspired progression markers.

The visual design draws from Altera AL's brand identity, using a deep green palette that evokes both technological sophistication and approachability. I incorporated a subtle motherboard pattern in the background to reinforce the laboratory/workshop atmosphere, suggesting a space where users can experiment and create. This theme extends through the interface elements, from the chat bubbles to the progress indicators, creating a cohesive environment that feels both professional and experimental.

Agent Management Page

After creating an agent, users can access a comprehensive management dashboard that displays key performance metrics and interaction history. The interface shows engagement stats like play count, interaction minutes, and user favorites - all with week-over-week growth indicators to track an agent's impact.

I designed two key features to help users maintain their agents: Save Files and Memories. Save Files act like game saves, allowing users to preserve different versions or "states" of their agent, while the Memories system provides access to past interactions and helps users understand how their agent evolves over time. The stats are displayed prominently with intuitive up/down indicators, making it easy to monitor an agent's performance at a glance.

The layout takes inspiration from game inventory systems, presenting agent saves as a grid of cards with clear timestamps. This familiar pattern makes it easy for users to navigate through their agent's history and manage different versions, while maintaining the laboratory/workshop aesthetic of the platform.

Agent Discovery Page

I designed a discovery system that makes it easy to explore and interact with AI agents, both from the lab and the community. Users can make their agents public, earning badges like "Highly Rated" and "Staff-Curated" based on performance and quality. The interface features recently played agents in "Dive Back In" for quick access, while "Community Favourites" showcases popular user-created agents.

The discover page takes inspiration from modern content platforms, presenting agents with rich preview cards that display key information like engagement stats and special tags. Lab-created preset agents are highlighted to help new users understand the platform's capabilities, while featuring standout community creations encourages users to experiment and share their own agents.

I used visual hierarchy and clear categorization to help users navigate between different types of agents - recently played, community favorites, and curated collections. Each agent card includes a tagline and visual preview, giving users a quick sense of the agent's personality and purpose before diving in.

Gallery