The AI Assistant is a conversational interface powered by generative AI that enhances productivity by helping users access insights, troubleshoot issues, and streamline workflows. It is integrated into Adobe Experience Platform and Adobe Journey Optimizer and can handle queries related to operational insights and product knowledge.
Operational Insights
AI Assistant offers detailed insights into various business objects such as audiences, schemas, datasets, and data flows. It can answer questions like "Which audiences have zero profiles?" or "Which source data flows are inactive?" This feature allows users to retrieve essential information about their data more efficiently.
Product Knowledge
AI Assistant can help users navigate and understand Adobe Experience Platform features by providing information on functionalities and workflows. It references Experience League documentation to ensure that answers are accurate and verifiable.
Interactive Features
- Discoverability: Users can explore the capabilities of AI Assistant through a light bulb icon, which suggests the supported questions and prompts that AI Assistant can handle.
- Operational Insights: The Assistant provides deep insights by querying specific business objects within the sandbox, allowing users to retrieve essential data like segments, attributes, and journeys.
- Multi-turn Conversations: Users can have natural, flowing conversations with the Assistant, allowing for follow-up questions and deeper exploration of topics.
- Verifiability: AI Assistant is built on trust, offering users a breakdown of how it arrives at its answers, including SQL queries and references to documentation, which helps validate the responses.
Roadmap and Future Innovations
The current AI Assistant supports both product knowledge and operational insights, with plans to evolve toward generative and predictive capabilities. The roadmap includes:
- Phase 1 (Assist): Focus on assisting users with knowledge and insights through the current feature set.
- Phase 2 (Generate): AI Assistant will help generate and automate processes like creating campaigns, copying segments, and more.
- Phase 3 (Predict and Recommend): AI Assistant will offer predictive insights and goal-driven recommendations, making it a more powerful tool for marketers and data teams.
Trust, Privacy, and Governance
Adobe has designed AI Assistant with strict privacy and governance controls. Key points include:
- Role-based access: Permissions for AI Assistant can be tailored based on user roles, ensuring that only authorized users access certain features.
- Data Privacy: Adobe ensures that data from different customers is never shared and does not use sensitive consumer data to train models.
- Verifiability: AI Assistant provides citations for all product knowledge queries and includes the underlying SQL queries for operational insights, ensuring transparency in how answers are derived.
Conclusion
The Adobe Real-Time CDP webinar provided an in-depth look at the power of AI Assistant in enhancing workflows within Adobe Experience Platform. With a focus on operational insights, product knowledge, and a roadmap for generative AI features, AI Assistant is set to become an integral part of Adobe’s customer experience solutions. The emphasis on privacy, security, and user control ensures that the tool is both powerful and trustworthy.