Amplitude's proactive product agent, Wave, leverages OpenAI models to slash product improvement time from roughly a month to just 1-2 days. The AI-driven tool has already demonstrated a 160% lift in engagement for Amplitude's own homepage.
TL;DR
- Wave, Amplitude's AI agent, uses OpenAI models to identify, prioritize, and implement product improvements.
- The tool has reduced improvement cycles from a month to 1-2 days, with some changes going live the same day.
- Wave's impact was evident in a 160% increase in agent engagement on Amplitude's homepage.
What happened
Amplitude's Wave is an AI agent that helps teams identify and implement product improvements. It integrates with OpenAI models, including GPT-6.1 Sol, GPT-6 Luna, and GPT-Image-2.5 Flare, to analyze data, generate solutions, and coordinate implementation.
Wave's workflow involves investigating product areas, developing solutions, and tracking outcomes. It can send tasks directly to coding agents like Codex for implementation.
In an internal test, Wave personalized suggestion cards on Amplitude's homepage, leading to a 160% increase in card-driven sends. The entire process, from discovery to execution, took just one day.
Why it matters
Wave significantly reduces the time and effort required to implement product improvements. This allows teams to focus on other priorities while Wave handles iterative enhancements.
The tool's ability to integrate with various OpenAI models enables it to balance quality, latency, and cost. This makes it a versatile solution for different types of tasks, from deep investigations to high-volume checks.
For developers and startups, Wave offers a powerful way to leverage AI for continuous product improvement. Investors may see this as a valuable addition to Amplitude's product suite, enhancing its competitive edge.
Key facts
- Wave is 80% powered by OpenAI models, including GPT-6.1 Sol, GPT-6 Luna, and GPT-Image-2.5 Flare.
- Wave can help teams ship product improvements in 1-2 days, down from roughly a month.
- In an internal test, Wave drove a 160% lift in card-driven sends on Amplitude's homepage.
- GPT-6.1 Sol scored 4.06 out of 5 in investigation tasks, with a median time of 54 minutes.
- GPT-6.1 Sol's data science agent scored 67.5 out of 100, with a mean model cost of $1.28 per dispatch.
- GPT-Image-2.5 Flare scored 80.3 out of 100 in generating solution wireframes.
- Among customer-launched experiments, one in four went live the same day Wave surfaced the opportunity.
- Wave is currently in beta, with customers across industries already using it.
Context
Amplitude has long been a leader in product analytics, helping teams understand user behavior and make data-driven decisions. With Wave, the company is extending its capabilities into proactive product improvement.
The integration of OpenAI models into Wave underscores the growing trend of AI-powered tools in the software development lifecycle. These tools can handle everything from code generation to testing and deployment.
As AI models continue to improve, the potential for autonomous agents like Wave to handle more complex tasks grows. This could lead to even faster development cycles and more iterative product improvements.
