API

Push your data — it appears in the workbench

One POST request uploads a CSV dataset to AB-Labz. The workbench mirrors exactly what you pushed. Delete the file any time — it disappears. Auto-cleanup after 14 days either way.

How it works

No sync protocol, no persistent connection. Just a file — pushed when you need it.

Push a CSV file

One POST request with your experiment dataset. AB-Labz validates format and encoding on upload — bad data is rejected immediately with a clear error.

The workbench mirrors it

The uploaded file appears immediately in the workbench under your experiment — ready for preprocessing and analysis. What you pushed is what you see.

Delete when done

Delete the file from the workbench — it's gone from AB-Labz storage immediately. Or do nothing: files auto-expire in 14 days. Your data doesn't linger.

Inside the workbench

API keys and uploaded datasets

The API settings page shows your active key alongside every dataset currently stored in AB-Labz — with file size, last update time, and days until auto-deletion.

workbench.ab-labz.com
AB-Labz API settings — API keys and uploaded datasets
Automate it

Schedule pushes from your side

The API itself is stateless — it just accepts a file. How often you push is entirely up to you. Set up a cron job or a pipeline trigger on your product's side, and AB-Labz will always have fresh data without anyone doing it manually.

  • Daily push — analysts open the workbench, data is already there
  • Hourly push during a running experiment
  • Event-triggered push from your pipeline
Full API reference

Python

# Push experiment dataset to AB-Labz
# Run this on a schedule — cron, Airflow, whatever you use
import requests
API_KEY = 'abn_xxxxxxxxxxxxx'
BASE_URL = 'https://workbench.ab-labz.com/api/v1'
EXP_ID = 'checkout_button_v2'
with open('dataset.csv', 'rb') as f:
r = requests.post(
f'{BASE_URL}/datasets/upload/',
params={'experiment_id': EXP_ID},
headers={
'Authorization': f'Bearer {API_KEY}',
'Content-Type': 'text/csv'
},
data=f
)
if r.status_code == 201:
print(f"Pushed: {r.json()['size_mb']} MB")

cURL

curl -X POST \
"https://workbench.ab-labz.com/api/v1/datasets/upload/
?experiment_id=checkout_button_v2" \
-H "Authorization: Bearer abn_xxxxxxxxxxxxx" \
-H "Content-Type: text/csv" \
--data-binary @dataset.csv
What unlocks

Live monitoring on the dashboard

Experiments delivered via API appear on the dashboard with real-time monitoring — updated automatically with every push you send.

  • SRM check — traffic split correctness verified on every push
  • Sample accumulation — see how far you are from the target sample size
  • Win probability forecast — Bayesian estimate updated with each new dataset
workbench.ab-labz.com
Active Experiments
AB Reduce steps in guest checkout flow
✓ SRM
ab-123xpt-ddw
CR Reg +5%
Success forecast 78%
Sample size 6 151 / 20 000
31% collected
Time left: 3 days
Data layer

One data mart — all experiments

The recommended pattern: build a single wide mart in your own infrastructure, then automate the slice-and-push step. AB-Labz receives clean per-experiment CSV files on a schedule — no manual work.

1
Build a single data mart

One wide table in your DWH — all experiments, all metrics, all segments, updated daily.

-- schema
dt | experiment_id | user_id | variant
| metric_1 | metric_2 | ...
| segment_1 | ...
2
Slice per experiment on a schedule

A nightly job queries the mart for each active experiment and exports the full dataset from day one to today.

# runs nightly
for exp_id in get_active():
df = query("SELECT * FROM mart
WHERE experiment_id=?", exp_id)
upload_to_ablabz(exp_id, df)
3
AB-Labz mirrors it instantly

Each push overwrites the previous file. AB-Labz always has the full up-to-date dataset — SRM, sample size, and forecast refresh automatically.

Single source of truth

All experiments share the same metric definitions — conversions, revenue, segments calculated once, consistently.

New experiments appear automatically

New experiment_id in the mart → next run pushes it. Just add the ID to the hypothesis card in AB-Labz.

Ratio metrics handled correctly

Store orders and revenue, not aov. The engine aggregates ratios correctly.

Data mart preparation guide

Available on Corporate plans

Start on Solo, upgrade to Corporate when you're ready to connect the API.

Try free See pricing