Autonomous buying has arrived, but most of it so far on CTV and the open web. Mobile in-app is next, and BidMachine's pilot is among the first to run exactly there. We invite you to test on live inventory at a contained scale, along with Amazon and other partners already in.
One backend, three protocols
The industry has not yet settled on a standard: AdCP, IAB Tech Lab's AAMP and Google's A2A are all in play. Our seller agent speaks all three through one backend, where a shared deal store, state machine and audit trail keep every agreement consistent, so terms negotiated over AdCP read the same as terms negotiated over A2A. Whichever protocol you have chosen, the pilot meets you on it.
"BidMachine was among the first to release a seller agent on both the AdCP and AAMP registries, breaking ground in a way rarely seen in the industry. Building it on Amazon Bedrock AgentCore helped them move faster, since running production AdCP agents on managed, secure infrastructure meant they could focus on the protocols instead of the plumbing. Having a working seller agent to test against full agentic media-buying workflows, alongside other customers, was especially valuable."
— Zelle Steyn, Principal Solutions Architect, Amazon Web Services
Discover, negotiate, transact
Starting today, your agent can do what a trader does: discover inventory, negotiate terms and create deals. If you don't run one of your own yet, you can test through ours. Every participant in the pilot is authenticated, with trust scoring in place as the pilot scales. Negotiations stay within set price boundaries and every action is logged, so you can push the system hard and see exactly what happened.
Early access
Pilot partners work directly with our engineering team and shape how these protocols get implemented before general availability.
This round is for demand partners, agency trading desks and independent developers.
To join, reach out to your BidMachine account manager or write to fernando.granata@bidmachine.io if we have not met yet.
_HighPerformer_HighPerformer%201.png)
