MercadoLibre
Q6The occupied grid is unchanged. The shape moves from the old Star read to Spear, because the load-bearing evidence is one engineering loop.
MercadoLibre’s sealed evidence is its own engineering loop. Agents independently reviewed more than 500,000 code submissions and migrated about 9,000 services in a year. Roughly 7% of AI-reviewed code is merged with no human approval, then still passes the normal tests. Q6 is occupied. The shape on the seal is Spear.
Trajectory: Q6 holds on the engineering loop.
Nothing quoted yet. A rate or a saving will sit here when it is on the record. It does not light the grid.
The lines this reading rests on. A date is the source date, when the page or the line names one.
Roughly 7% of AI-reviewed code is now merged autonomously – accepted into our codebase without requiring human approval – before passing through our standard testing and deployment procedures.
As engagement rises, our assistant is resolving a growing share of interactions: in Q2'26, more than half of the requests on our Help Portal were solved by the assistant without any human intervention.
Mercado Ads' AI Advisor reaches sellers on WhatsApp, analyzes campaigns in real time and delivers fully automated budget and ROAS recommendations with zero human intervention. It has scaled from small pilots to tens of thousands of sellers per month, and one in three sellers who engage with it go on to make a change to a campaign directly within the conversation.