How do I deploy Reifier?
There are multiple options to deploy Reifier as per your data volumes, existing(or none) big data infrastructure and choice of cloud provider
- As a managed service on public cloud like AWS, Microsoft, IBM
- As a library on a Spark cloud service like Databricks Cloud, AWS EMR, Qubole, Microsoft HDInsights
- On premise within customer data centre
- Custom tailored deployment
How long does it take to train Reifier?
Reifier is typically setup within 2-3 hours on your data
I dont have a big data environment, can I still use Reifier?
Absolutely. A single machine Reifier instance is good for a couple of million records. You could also use Reifier over a cloud managed Spark environment like Amazon EMR.
Can I use Reifier for multi domain MDM?
Yes, of course. We like variety 🙂
How many training samples are needed?
It depends on the data, the noise and the number of fields, typically the interactive learner converges well for about 20-30 training pairs
I do not have a background on big data or machine learning. Can I still use Reifier?
Absolutely. The AI algorithms run in the background, the user does not need to worry about them
I want to use Reifier for real time matching. Do you support that?
Sure, real time matching is very much supported.
Posted on August 19th, 2016