Acera Labs
BRING YOUR OWN MODEL

Plug in the model your team already built.

Your model. Our decision layer.

Acera Labs reads your saturation curves from any source. Recast, Robyn, Northbeam, Google Meridian, Mutinex, custom Python notebooks. Eight months of model investment should compound, not get rebuilt.

Australian-hosted · Opt-in privacy by default · ABN 56 506 585 125

SUPPORTED MODELS

Eight model sources, one decision layer.

Connect your existing model. No rebuild required. Acera Labs reads the saturation curves overnight and prepares ranked decisions by morning.

RC
Recast
JSON / native API
RB
Robyn
R output files
GM
Google Meridian
JSON / BigQuery
NB
Northbeam
CSV / data share
MX
Mutinex
Data export
PY
Custom Python notebook
Any schema
PR
Prescient AI
Export API
TW
Triple Whale
Data share
HOW IT WORKS

Connect once. The agent reads it every night.

Step 01
Connect your model output
Deliver your model in any supported format. Recast JSON, Robyn R files, Northbeam CSV exports, Google Meridian BigQuery output, or a custom Python notebook with a defined output contract. A one-time channel mapping confirms the schema.
Step 02
Acera Labs reads your saturation curves
Every overnight run, the agent reads your model output and marks each channel's position on its saturation curve. Current ROI index, headroom remaining, and diminishing-returns position are extracted and stored alongside live spend data.
Step 03
The agent uses your model in the decision queue
Ranked decisions arrive in the queue each morning with your model's evidence attached. Each packet shows which channel is below saturation, how much headroom remains, and the projected outcome of acting versus waiting. Nothing activates without your approval.

Your model. Our decision layer.

Upload once. The agent reads your saturation curves overnight and delivers ranked decisions by 7 a.m.