February 17, 2015 Comments Off on Fast Company names Poshly one of the Top 10 Most Innovative Companies in Big Data
I blogged about Poshly some time ago. One reason that I invested in them was that they’re a great example of employing Big Data to answer real-world questions, rather than just vacuuming up a bunch of information and trying to find a use for it.
Bradley Falk, Poshly’s CTO and co-founder states:
The great thing about beauty and personal care data is discovering how unique everyone is. We can create a portrait of a user in near realtime and discover how the small details can vary so much. We can react to trends, interests and sentiment to create value for both the consumer and the industry while protecting the user’s personal information.
As I’ve watched Poshly’s meteoric growth, I’ve been interested about the approach they would follow to maintain scalability. According to Matthew Drescher, Poshly’s Head of Data Engineering:
We are aggressively utilizing high performance, distributed in-memory computing techniques to vectorize our data, perform in-place analytics, and paint a landscape of insights for our customers to enjoy.
With the quality of data Poshly gathers, it is possible to take a very geometric approach to generating insights. It’s less like scraping through a haystack in search for a diamond than it is trying to realize the maximum realistic photo resolution.
If you’re interested in all things Big Data, stay tuned for a series of blog posts I’ll be writing on critical algorithms that should be part of your toolkit.
August 31, 2014 Comments Off on Poshly: a great practical example of Big Data in action
With any new innovation, hype can often outstrip results, especially in the early days. A few years back, we saw this with Service Oriented Architecture (SOA): it had lots of promise, but there were relatively few examples of successful implementations. Nowadays, SOAP and REST services – supported by the principles of service orientation – are the primary techniques that distributed applications use to communicate. This has led to all sorts of innovative solutions, especially when pairing these services with mobile devices.
The same things are happening in Big Data: you hear about it all the time, but it’s natural to wonder how it’s being used to add value. Unsurprisingly, new technologies are often viewed as solutions in search of problems, and this is particularly relevant for Big Data since it’s such an all-encompassing discipline.
For Big Data, it’s always useful to look for practical applications: first define the problem, and then use Big Data to supply the solution. Poshly is a textbook example of what I mean: Big Data technologies and practices are being applied to meaningful problems, thereby helping customers answer questions that were very difficult to resolve prior to these advances.
Disclosure: I’m an investor in Poshly through my participation in Astia Angels, an organization I encourage you to check out. Poshly recently closed a $1.5MM investment, which you can read about here at TechCrunch.
Poshly’s website offers a variety of beauty product giveaways which consumers can compete to win by answering personal questions about their beauty routines, habits, interests, and more. The data these questions generate is highly personalized, but only shared with Poshly’s brand customers after being anonymized – meaning users’ personally identifiable information is removed, like their name, email or address. This “hyper-personal data,” as CEO Doreen Bloch calls it, helps brands better understand their customer base in general, or influence larger decisions, like what retail channel to roll out to next, for example.
I’m increasingly learning about dynamic startups like Poshly that are finding realistic uses for Big Data. As time goes by, I suspect that eventually we’ll stop using the term Big Data, and depict it instead as just plain “data” as we portray the exciting ways that information is being put to work.
November 25, 2013 Comments Off on Hadoop Buyers Guide is now available
Choosing a Hadoop platform can be confusing: there are several great alternatives on the market right now. Some of these offerings require you to handle all aspects of installation, configuration, and administration on your own, while others deliver a more comprehensive, innovative, and integrated solution yet are still faithful to Hadoop’s open source heritage.
I recently put together a concise eBook that you can use to help get a better understanding of your options. You can view the guide here.
October 21, 2013 Comments Off on eSeminar on “Using the Cloud to Support Today’s Growing Analytics Requirements”
I recently conducted an event for IBM & Ziff Davis that talked about the inter-relationships among Big Data, cloud computing, and analytics. From the abstract:
Analytics is being driven deeper into most organizations as more business unit managers and employees need information to make smarter, faster business decisions. At the same time, many companies are relying on the analysis of much larger volumes of data to make such critical decisions. This eSeminar will discuss how a smart cloud infrastructure can help meet today’s growing analytics requirements, while accommodating the larger volumes of data involved in decision-making processes.
You can view the eSeminar here