What’s in store for data analytics in 2015 and beyond

Future Gazing

Since the last couple of years, we have experienced revolutionary changes in the data analytics arena.  Big data became main stream and Data Scientist got termed as the sexiest job.  Data explosion has come with its own fair share of challenges and opportunities.  As the businesses are looking forward to monetize the data, they are grappling with finding suitable business use cases to leverage and execute on the new data platform.  Added to this is the trouble of finding skilled workforce to address the burgeoning demand.

Recent Alteryx research report has found that 72% of the business leaders are not satisfied with the speed at which insights are derived from data and nine out of ten dissatisfied stakeholders blame others on the inability to blend data from various data sources.

What to expect?

Segmentation of Vendors – Industry is fast gyrating towards three categories of vendors who are providing technologies that assist the data analytics life cycle viz. Data Platform providers like Cloudera, MapR, Hortonworks etc., Data Wranglers such as DataRPM, Tamr etc. and Guided Data Discovery providers like Alteryx, Datameer, Databricks etc.  Adding interesting twist to this mix is the new age big data visualization vendors like Platfora, Tableau etc.

It’s something of an irony if you find striking similarity to the data world where we used to live in until recently which was dominated by Oracle, DB2, MS SQL Server etc., as Data Storage providers, Informatica, Datastage, Abinitio etc., as ETL providers and Cognos, Business Objects etc., as the vendors for BI / visualization.

Few other distinct trends which we would emerge in days to come would be:

  • Machine Learning would become main stream
  • Deep learning could fast replace machine based learning technologies
  • Most of the predictive modeling and data science could become code-less or automated
  • Everyone would become data analyst or at least would be able to do sophisticated analysis through the power of these new found tool sets
  • Data blending would become main stream and rapidly get commoditized – a real threat to Big Data based analytics
  • Adoption of real time streaming and analytics would be higher – expect Apache Storm and Spark to be heard more often than before during your discussions
  • Signals based approach would become more prevalent than the atypical use case driven analytics approach where advanced analytics would be driven by a library of domain based signals.  We already see companies like Opera Solutions, Platfora, IBM’s Watson etc., taking lead in this space.
  • AI / VR could become an integral part of the strategy for the forward looking organizations
  • Graph databases would be increasingly used as the new destination repository for guided discovery / analytics
  • CIOs would have minimal say on the choice of analytics / tools / methodologies which would be driven by the business through the newly created roles like CDO (Chief Data / Digital Officers) or CAO (Chief Analytics Officers)

How many of these trends have you seen already kicking in?  Or have I missed out something more critical and significant?  What do you think?

Please share your thoughts.

Let’s ask bigger questions…

Image Credit – Morguefile

Why Microsoft could become a force to reckon with in Machine Learning

Microsoft has gotten into a great course correction since Satya Nadella took over as the CEO. In its effort to transform itself into a platform company per the vision of the new CEO, Microsoft has been truly taking some critical steps in the Big Data analytics space.  Since the launch of its Azure Machine Learning platform, Microsoft has been quietly focussing on building the foundational blocks and consolidating the Azure MLs adoption.  It’s easy drag, drop and predict approach on a cloud based Machine Learning platform has won a solid following.  It has also been helping developers to jump start Machine Learning through its Microsoft Virtual Academy led courses such as :

Microsoft has also recently made a couple of rather significant acquisitions to bolster its case further.  It’s acquisition of Equivio, a Text Analysis / Machine Learning based eDiscovery / compliance vendor whose main product is Zoom, a court approved machine learning platform.  The deal is stated to be around $200 Million.  Microsoft is reportedly planning to utilize Equivio’s machine learning technology to further improve its Office 365’s eDiscovery and information governance capabilities.  It would help quickly search and find relevant information from the unstructured data present in documents.

Yet another recent acquisition had been Revolution Analytics – a commercial software and services provider for R, the world’s most widely used programming language for statistical computing and predictive analytics.   It has also extended its support for open source communities by adopting Linux and also partnered with Hortonworks, a open source Hadoop distribution vendor to extend Big Data to the enterprise through its Azure HDInsight Big Data platform offering.  Microsoft had also open sourced its REEF to provide a big data analytics framework for YARN.

Microsoft is also facing off with IBM Watson through its ease of use Machine Learning approach.  It’s attacking IBM Watson’s biggest issue – the steep learning curve by providing an easy to use and adapt data model / APIs.  In its recent post, ZDNet reports that Microsoft’s goal for Azure Machine Learning is to develop data models that can be plugged directly into apps that will take that data, analyze and query it, and turn it into information that will be greater than the sum of its parts for a user.  Point in case its prebuilt data model for Microsoft Band which is its first fitness device.  The device monitored data is used for measuring, analyzing the health parameters.  It functions as a hub for data analysis of all the data collected through the device monitor.

Interesting times ahead for Microsoft since it seems to have placed its bets big on Analytics, Big Data, Cloud & Machine Learning in a big way.

Free Big Data and Data Science Ebooks

A bunch of free Big Data and Data Science Ebooks which you should find useful.  You could download them for free using your Amazon Kindle.  If you don’t have a Kindle device, you could download it using a Kindle app for Android / iPhone / Mac / Windows.

I’d wish to keep this list active.  Do post your comments if you find any new Ebook that is free and covers either Data Science or Big Data.
Hope you enjoy reading them.

 

TDWI Chapter (India) Event on Big Data Technologies & Perspectives

TDWI Chapter (India) event (FREE) on Big Data in Hyderabad on 23rd July 2011 hosted by Infosys

Agenda :

09:00 10:00 Registration and Coffee
10:00 10:15 Welcome Address – Srinivas Prabhala, CTO, Financial Services, Infosys
10:15 10:45 Data Deluge: Opportunities and challenges N T Arun Kumar, Managing Director, UBS
10:45 11:45 Technical Challenges of Handling big data Ramesh Bhashyam, CTO, Teradata (India R&D)
11:45 12:45 Hadoop and No SQL Approaches Nagaraj Kulkarni, Chief Architect, Compugence
12:45 14:00 Lunch Break
14:00 14:45 Data Analysis through DW Appliances Satyanarayana Ranganatha, Sr. Manager, EMC
14:45 15:30 Innovations in Analytical Databases Vivek Bhatnagar, Country Director, Ingres
03:30 03:45 Tea Break
03:45 05:15 Panel: Some Perspectives on Big Data

  • Sreenivas Simhadri, General Manager, Microsoft
  • Ajit Nema, Director, Deloitte
  • Suresh Donepudi, CIO, ITC Bhadrachalam
  • Gurpreet Singh, CEO, DEFTeam
  • KVSN Sastry, Principal Architect, Infosys

5:15 5:30 Closing Keynote – Subbu Venkatraman, Vice President, CA

Registration are mandatory to attend the event.  Register using http://tdwichapters.org/pages/india/event-registration.aspx

For Clarification send an email to tdwi.india@gmail.com