ABK Assist

Summary & Key Components

Overview:

Activision/Blizzard ABK Assist is an internal product built to deliver contextual AI experiences across the platform, serving front-line workers in customer service, IT, and HR
 

The product was not being utilized in any meaningful or product way because of mixed messaging, structure of work process and non-collaboration from dependent teams, plus information and requests were dense and chaotic 
 

The platform team was overworked, burned out and barely addressing flaming hot immediate issues 

Objectives

• Unblock the development and integration of AI tools in the ABK Assist platform

• Improve collaboration for scaled implementation interdivisionally 

• Build a healthy team culture—especially across the messy seams of horizontal product work     

Obstacles

Communication was siloed and transactional and directives were given in dense information without context

No program manager meant constant urgency and a lack of prioritization. Couple that with embedded designers that were excluded from decisions and disconnected from the core team.

​

This lead to lack of structured process caused confusion and disengagement. 

Solutions

I led the ABK Assist AI Design team to: 

• Implement structure, clarity to define process

• Shift from isolated delivery to inclusive, cross-functional and cross divisional collaboration that improved clarity, team culture, and partner engagement 

• Rebuild trust with internal and external design teams through clear engagement practices, task and goal setting

Re-engaged embedded designers and scaled inclusive meeting rituals which boosted moral and set a rhythm to the pace of each sprint.

I also championed smarter project planning through a team-led capacity tracking, transforming platform AI implementation. This helped to redistribute the work-flow and deadlines more equitably with staggered teams. 

​

I built and implemented:

I took a servant leadership approach: No sweeping directives, just attentive scaffolding to
facilitate the group to thrive. I also developed a lightweight capacity planning and sizing model  
to determine and manage future resource needs by matching projected demand with the team's capacity. 

​

• Developed and launched the ABK Assist Context Menu

• Launched the ABK Assist Summarization Skill

• Intelligence to improve the agent experience Developed ABK Assist AI Skills and Predictive

• and service to our customers

Designing AI to Scale - The Changes

ABK Assist Contextual Menu

• Designed to scale with evolving customer needs.

• This MVP feature allowed customer service agents to summarize, create, and edit written text using generative AI. 

• I guided the designer to think beyond near-term usefulness—making the interaction model extensible as we introduced additional features like:

  • Inline actions

  • More fine-grained recommendations

  • Context-sensitive editing

ABK Assist Skills and Predictive Intelligence 

• Patterns designed with purpose.

• This MVP feature allowed customer service agents to summarize, create, and edit written text using generative AI. 

• I guided the designer to think beyond near-term usefulness—making the interaction model
  expandable as we introduced additional
  features like:

  • Inline actions

  • More fine-grained recommendations

  • Context-sensitive editing

ABK Assist Contextual Menu

• Fast clarity for complex cases.

• Implemented throughout the platform, this skill allowed customer service agents to instantly grasp the context of support cases - summarizing customer intent and previous assistance. 

• My design leadership concentrated on clarity, speed, and user confidence.

Outcomes and Impact

ABK Assist Contextual Menu

The Transformation Was Real

• Healthier work management: 

  • With structure, intention and process flow - Designers reported their work was more focused and manageable.

• Increased cross-team efficiency: 

  • Embedded designers became involved again and were actual contributors.

• Thoughtful enablement: 

  • We started creating enablement content WITH our
    partners rather than FOR them, leading to more relevant tools that addressed actual needs.

• More intuitive AI tools: 

  • With better-defined roles, improved cross-team
    alignment, and tighter design governance, we shipped scalable AI features that supported customer service agents in real-time.

The teams felt more empowered, connected, and confident - indications of a culture in which servant leadership is working and nurtured.

​

Operating as a Team

How we operated as a team was immediately noticeable internally and externally.