Back to Projects

Generative Audiences

AI-powered natural language interface for building customer segments

Segment

Overview

Generative Audiences transforms how marketers build customer segments by allowing them to describe their target audience in plain English. Instead of navigating complex UI filters and dropdown menus, users simply describe who they want to reach, and AI automatically generates the precise audience conditions in Segment.

This innovation dramatically reduces the technical barrier to audience creation, making powerful segmentation accessible to marketers of all skill levels while significantly reducing time-to-market for campaigns.

Key Features

Natural Language Processing

Users describe their audience in plain text, such as "Users who purchased in the last 30 days but haven't opened our emails" and the system interprets the intent and context.

Workspace-Aware AI

The LLM receives full workspace context including available events, properties, and traits, ensuring generated conditions are valid and use actual data from the customer's Segment workspace.

Intelligent Condition Builder

AI automatically translates natural language into structured audience conditions, handling complex logic, time-based rules, and nested conditions without manual configuration.

Impact

  • Reduced average audience creation time from 15 minutes to under 2 minutes
  • Enabled non-technical marketers to build complex segments independently
  • Increased audience creation volume by 40% among existing users
  • Reduced support tickets related to audience building by 35%
  • Improved accuracy of audience conditions with AI validation

My Role

As Staff Product Manager, I led the product vision and execution of Generative Audiences from concept to launch. Key responsibilities included:

  • Defined product vision combining LLM capabilities with Segment's audience building platform
  • Conducted user research to understand pain points in traditional audience creation workflows
  • Partnered with ML engineers to design the prompt engineering and context injection strategy
  • Collaborated with design to create intuitive natural language input experiences
  • Established success metrics and instrumentation to measure AI accuracy and user adoption
  • Led go-to-market strategy and positioned as key differentiator in competitive landscape
View More Projects