Nexios provide variety machine learning APIs. Main use cases are classification, regression, forecasting and anomaly detection. So it's more an ML API or MLaaS than a machine learning platform.

You code against their API or API libraries. So it's a black box model approach.

client.datasets.create('coffee-consupmption', data)

# or

with open('cups_of_coffee.csv') as f:
	client.datasets.create('coffee-consumption', f)

start_date = datetime.datetime.date(2017, 12, 1)
end_date = datetime.datetime.date(2017, 12, 31)

client.sessions.create_forecast(
    'coffee-consumption',
    'cups',
    start_date,
    end_date,
    result_interval=TimeInterval.Day
)

import_response = client.imports.import_from_s3(
    'test-python-import', 
    'sample-data', 
    'some-file.csv', 
    'us-east-1'
)

Sources

• Nexosis