Perceptor: Advanced Cloud Machine-Learning Platform

A cloud machine-learning platform to create, train and deploy models.

Technologies used

Python, Kubernetes, Spark

Installation process

Perceptor is an Integration Platform as a Service (iPaaS) solution.

Why you need Perceptor

Create, Train, and Deploy ML Models on Cloud

Perceptor is a scalable platform that allows users to create highly customizable ML pipelines by leveraging all kinds of data, including structured and un-structured.

Create ML Pipelines

Create ML pipelines with/without Data Science, ML or statistics background.

Customizable

A flexible solution to create models by different level of customizations from UI driven solution with minimal manual intervention to complex programming.

All-in-One Solution

Use any type of data including structured and un-structured and tackle any data problem, including supervised, unsupervised and out-of-box solutions.

Scalable

In case of high-volume tables, Perceptor can be scaled on demand if required

Why Perceptor

Perceptor has the following features and advantages:

UI: Single-pane-of-glass view for all types of datasets and users.
Scalable Infrastructure: Having capability to scale up and down as per data / demand with cloud based processing engine in place.
Comparison between models:: Perceptor allows you to monitor accuracies and performance of different ML models along with graphs for different iterations and across different time periods..
Simplifies ETL process: Perceptor provide different connectors for all types of data sources including excel files, hadoop clusters, data bases and cloud platforms.
Future Versions:

Unsupervised Learning algorithms will be introduced.

More Deep Learning algorithms will be introduced.

NLP and Computer Vision related models will be added for text, images, speech related data modelling.

Other Out-of-Box solutions will be added like POS tagging, NER, Topic Modelling, Image Augmentation, etc.

Perceptor