Big-data happens to other people
Last summer, we created our most ambitious startup to date. Today, we are lifting the lid on how we plan on fundamentally changing the way people engage with data.
A look back
Redsift was born of a personal frustration of mine. I have been responsible for creating and commercialising non-trivial technology for ~15 years. If you think back to your first iPhone and the excitement of mass market App adoption, most likely your first few downloads included one I had created.
Shazam was the sort of product that the technologist in me truly loved, a simple idea coupled with fiendishly complex technology to create something that ultimately felt straightforward yet magical at the same time. Over the years, the march of technology has allowed us to make step changes in what is possible. Today’s state of the art cognitive computation models allow machines to process data and hence understand and make decisions about aspects of our lives in ways that will profoundly impact our relationship with computing.
Yet, through all of this, I personally can’t leverage much of what is now possible to simplify and automate my life. I am still using the same archaic menu of products everyone else is and expending energy on tasks I know computers would be better at.
Necessary but not sufficient
This insight in itself is not revolutionary. We know there is an oft discussed and fictionalised AI Singularity somewhere on the horizon. I believe I will see this in my lifetime but that puts me in the optimistic camp both with regards to the rate of technology advancements and my longevity. However, I don’t believe the building blocks we have today are sufficient to form the precursors of this true emergent AI and it is not entirely apparent what needs to happen for this idea to become a reality.
What I am more excited about is the here and now. We are not leveraging the advances already made over the last decade to their potential. As we look around to the companies that really embraced and exploited Machine Learning, Natural Language Processing and related technologies, we end up in the same cluster of vertically focussed solutions. As technology creators and innovators, is optimising ad-tech really the most impactful use of this class of technologies? I am being a little facetious, there are a number of really interesting companies out there looking to exploit cognitive computing in new and innovative ways. Ultimately though, the problem space is vast. As I look at the data challenges facing me as an individual and then scale my problems across the organisations I have been a part of, I see hundreds of micro-vertical opportunities. Lots of places where I need intelligent agents that can access my digital data streams and archives with the objective of monitoring, summarising and automating tasks that are currently offloaded to my own limited cognitive abilities. What I really need is a cloud where these digital robots can securely access my data and do their magic.
Helping humans
With Redsift, we created this cloud. Most of us are familiar with clouds that connect us and store our data. Yes most of these clouds amount to filing cabinets in the internet. They provide fundamentally important functionality, but they do not let us compute i.e. do work on the data that is in them. Redsift is a new style of cloud, it is Platform as a Service for your data. It provides a home for a new kind of App, we call it a Sift. You can think of a Sift as a always-on digital agent that can stream data from your silos and can present the things you need to know back to you. We are currently working on Sifts that we believe will give our users a whole new class of insights from their data but ultimately, of all the problems we need to solve we will only be able to address a few. That is why Redsift is an open platform; if you want to create your own Sifts for your own data, you can.
All the parts
Our grand vision starts with a single step. We need to combine our architecture with algorithms and most crucially, with data that we can extract insight from. We started with email as this is a large lake of data that every individual and organisation has and struggles with today. Our email Sifts provide automation and insight for your overflowing inbox but this is just the start. Over the coming months we will be showing what Redsift means when it is mixed with your Slack channels or IoT devices. If email is today’s data lake, your virtual and physical Bots are probably tomorrows and we want to solve your data problems across every channel.
We will be sharing the Sifts we are working on and talking about how we put them together. At launch, we will open source these Sifts so anyone who wants to deep dive into our platform can look under hood or even fork them and make them fit your needs. Our platform is currently in internal beta, if you would like an invite to our early access program, register here or email us with your ideas and we will bump you to the top of the list.
