Image: Amplitude
Why it mattersCoding agents made it cheap to build. The slow part is now deciding what to build. An agent that reads product analytics is a different tool from one that writes code.
Coding agents made it cheap to write a feature. Picking the right feature is now the slow part. Amplitude says Wave is for that. The company launched Wave on 12 November 2026. Amplitude calls Wave a proactive product agent. It looks at how customers use the product. It reads the product code. It then writes up what the team should build next. It measures whether each change made things better for users.
Amplitude's launch post gives three customer numbers. Each one is attributed to the company.
What Wave does
Wave sits beside the team's analytics. It links session replay, web performance and user behaviour in one view. Darshil Gandhi, Director of Product Marketing at Amplitude, says the agent "recommends what is worth building". He adds that it "measures whether new changes make your product better". The agent writes each suggestion into a workspace Amplitude calls an Opportunity. The team reads the Opportunity, builds the change with a coding agent, and Wave then reports whether users responded.
Amplitude does not name the model or the model provider behind Wave in the launch post. Teams who need that detail should ask the vendor.
Who the launch partners are
Amplitude names The Economist, PetScreening, PixAI, Siemens, OneClickCode, PrizePicks, Rocket and DigitalOcean as launch customers.
PixAI, an AI image product, reports a 76 percent increase in Missions opens after acting on Wave's suggestions. It also reports close to 10 percent higher publishing and a 3 percent lift in Day 1 retention. Amplitude's own internal test reports 160 percent more suggestion sends in one feature.
DigitalOcean reports 55 opportunities found in six weeks. One DigitalOcean lead told Amplitude that "my time spent writing separate product requirements documents (PRDs) went to almost zero". Every figure above is Amplitude's, or a customer's. The launch post is the source.
Where it fits in a working team
Jacob Newman, a Principal Product Manager at Amplitude, wrote a separate post about his own workflow. He uses Wave to watch replays. He proposes a change inside an Opportunity. He then measures the result after release. He says the method is slow where it should be slow: a real experiment can still take months to read out.
A team that already pays Amplitude can test Wave against its own product. The hook is one view across behaviour, code and business metrics. The person deciding what to build sees all three together in one place. Amplitude's post does not state a price. Teams who want to try Wave should ask their account manager for access and pricing.
The launch post names a category Amplitude calls proactive product agents. A buyer comparing Wave with any similar product should ask the same questions Amplitude's own post answers: which customers used it, what number went up, and how that number was measured.
Source
Amplitude Wave launch post by Darshil Gandhi. Jacob Newman on using Wave.
This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.
Get AI News in your inbox
New developer tools, model and agent releases, and how teams are actually using them to release software. Short, and only when there is something worth reading.
