Start-up data player Iguazio is named in reference to – and with reverence for – the Igauzu Falls that lie along the Brazilian-Argentian border.
Meaning: big water | iguazio = iguazu+io | Pronounced: ee-gwa-zee-o
The company specializes in what it calls continuous (data) analytics. So this technology performs big data number-crunching, driven by a software engine designed to ‘drink from the firehose’ of constantly produced data emanating from devices that are always on.
The link to the always-on nature of the sensors, monitors and systems in the Internet of Things (IoT) is not hard to make.
Data is food, an analogy
What Iguazio has done is to engineer its data analytics engine with a big mouth and lots of differently shaped teeth. In this way, it has massive data ingestion capabilities, a complex data pipeline to feed from (food coming from different directions in different shapes and sizes, if you will run with the analogy), as well as a huge data integration task to deal with most of the time (this is chewing and food mastication before swallowing, obviously).
While our food and eating analogy might (arguably) be a much better way to explain what Iguazio does, the waterfall analogy obviously works well, too. And besides, if the company had used the massive food theme, it would have had to call itself Data-FatBoy… and that would just be silly.
The company says its real-time continuous analytics platform reduces time-to-insight from hours to seconds, eliminating data pipeline complexities, while seamlessly integrating with Apache Spark and Kubernetes. We know that Iguazio ingests, enriches, analyzes and serves data from one single software base – or a ‘unified platform’, as marketing people like to say.
New data + historical context
“To remain competitive, enterprises need to act upon business insights that are continuously generated from data collected from multiple sources and enriched with historical context,” says Asaf Somekh, Iguazio CEO.
“Iguazio works with this constant flow of fresh data from streaming sources, blends it with previous insights and historical data from multiple repositories and generates fresh insights that can be viewed in an interactive dashboard for actions and insights.”
Somekh insists that continuous analytics allows for information to be generated with sub-second granularity. It is, essentially, a complete rethink of the traditional data pipeline, where data is collected into static files and logs and processed later in a multi-stage approach.
Read more: The rise of the IoT ‘megatrends’
Why stateless analytics matters
Iguazio integrates with the open source frameworks of Spark and Kubernetes to help create ‘stateless analytics services’ and data processing tasks. This matters, because stateless (as opposed to stateful) computer procedures describe those functions where the machine is specifically programmed not to remember preceding user interactions, related data sets or other outside element data events.
The use of stateless analytics services can be useful in IoT data analytics, when we want to segment out data blocks to examine individual components of a complete architecture, without necessarily being tied to their relationships with the total universe of data, devices and management layers within which they exist.
As WhatIs from TechTarget very beautifully puts it, “Most computers, human beings and elephants are stateful.” But sometimes the IoT shouldn’t be like any of those things in order to do its job well.