
How Amplitude ships product improvements in 1–2 days with OpenAI

Amplitude has built a proactive product agent called Wave that helps teams turn product analytics into shipped improvements. Amplitude describes Wave as roughly 80% powered by OpenAI models, including GPT-6.1 Sol, GPT-6 Luna, and GPT-Image-2.5 Flare, and estimates that Wave can help teams ship product improvements in 1–2 days, down from roughly a month. Teams give Wave a product area plus the goals and metrics they want to improve; it then combines Amplitude data, session replays, customer feedback, and code to identify and prioritize opportunities, develop solutions, coordinate implementation with people and coding agents, track outcomes, and feed learnings into the next round.
Different OpenAI models handle different parts of that workflow. GPT-6.1 Sol supports the opportunity manager and deep investigations into code and data. GPT-6 Luna handles frequent, lower-cost tasks such as scanning code and running small evaluation checks. GPT-Image-2.5 Flare generates product mockups using screenshots from session replays as visual references. When an opportunity is ready to build, teams can send it directly from Wave to coding agents, including Codex, with a technical plan and acceptance criteria to guide implementation.
Amplitude tested Wave on its own homepage, where generic suggestion cards were meant to help users start a conversation with an AI agent. In a sample of 100 session replays, 80 users typed their own questions and only 12 selected a suggestion. Wave proposed personalizing the cards using recent charts, dashboards, or the page a user had come from; it illustrated the change and produced a technical plan that preserved free typing and showed generic suggestions when personalization context was unavailable. After a brief review, the team handed the work to a coding agent and A/B tested the change. In the readout Amplitude shared, messages sent through homepage suggestion cards increased 160% relative to control after one day.
Amplitude chooses models per task by evaluating quality, latency, and cost. For Wave‘s Opportunity Investigator, GPT-6.1 Sol scored 4.06 out of 5 overall, compared with 3.62, 3.52, and 2.94 for three alternative models; it led on all five investigation criteria and had the fastest median investigation time, 54 minutes. In a separate evaluation of Wave‘s Data Science Agent, GPT-6.1 Sol scored 67.5 out of 100, compared with 69.3 for another leading model, at a mean model cost of $1.28 versus $2.63 per dispatch, and it scored higher on tool-contract compliance, estimation and uncertainty, and evidence grounding. For high-volume work, Amplitude uses GPT-6 Luna to keep latency and cost low, reserving more expensive reasoning for tasks that need it most. For wireframes and mockups, OpenAI‘s image generation model scored 80.3 out of 100 for solution wireframes, compared with 63.9 for another image generation model, based on communication quality, product fidelity, instruction adherence, text and UI legibility, and visual coherence across 15 opportunities.
Customers in beta can control the level of automation. Most teams send bug fixes straight to coding agents, while new features and bigger bets get a human review. Among customer-launched experiments, flags, and guides, one in four went live the same day Wave surfaced the opportunity, half in under four days, and two in three within a week; the article notes these figures are interpolated from supplied milestones and the curve is illustrative, not individual observations.
The article closes with quotes from Amplitude CEO Spenser Skates about AI finding useful work and about balancing quality, latency, and inference cost across the product. Skates says the ambition with OpenAI is to make “self-improving products” the way companies build software; as OpenAI‘s models get better at reasoning and coding, teams should be able to delegate more investigation and implementation work to agents. Amplitude reports that agents are already finding and helping deliver improvements in its own product while the team works on other priorities.

