What AzureCon Means For Businesses
It sometimes feels like the latest technological discoveries are just around the corner – however they are only for the elite. Those willing to spend, those with some special knowledge. However AzureCon showed Microsoft in a different light to that stereotype.
It all started with me connecting my Surface Pro 3 to my TV, in preparation of 3 keynote speeches at AzureCon (Microsoft’s “this is what you can do, and this is what’s new” session). Not being a ‘techie’, the following breakout sessions would have only confused me, so this blog is from a business point of view.
It was, without doubt, an enlightening 3 hours! Microsoft demonstrated how the Cloud is now progressing at an extremely aggressive pace, with seemingly limitless potential. This stems from an atmosphere that is solely focussed on innovation and destroying boundaries.
There were a few standouts for me from the keynotes, which I thought I’d share in case you missed the sessions:
The SQL stuff made me think “we’re not in Kansas any more Dorothy”.
Watching one person stand up a SQL data warehouse would be very impressive, to watch someone do it during a live demo, more impressive again. But to do both of those things, in less than 5 minutes was quite frankly, staggering. The whole ‘fail fast, fail cheap’ mantra seems evident with Azure.
This completely changes the potential of a business – you can just click a button, move a slider and there you have it. That’s how you outpace your competition!
Data warehouses I get. Data Lakes has moved things on far enough that the idea wouldn’t look out of place being used by Patrick Stewart and the guy with the beard on Star Trek: The Next Generation.
The amount of data we’ll now be able to process, with (relative) ease is staggering. It’s commonly expressed that there are more stars in the universe than there are grains of sand on Earth. It won’t be long before similar statements will be made about Data Lakes… “There’s more data in this Lake than the whole world created before 1992” – ish… When Microsoft say unlimited, it would seem they mean it.
That brings me nicely on to data analytics…
Cortana Analytics Suite is much hyped right now. Rather than one product, it’s more like Office 365 – a selection of tools and products that work well together, but equally standalone. For a bit of context, I’ve never been a Big Data guy. I’ve worked for the past few years on Fast Data – narrow, deep and specific interrogation of large data sets to hit what’s needed rather than extrapolation of patterns within rich data sets.
Primarily, that’s because good Big Data seemed beyond the reach of most organisations… without the data processing power, the data science to interrogate and query the data you had a massive dataset and little potential.
Again, Microsoft seems to have moved this on. The stand-out thought I had watching the Big Data demos yesterday was that Big Data, useful Big Data, is achievable for organisations without the manpower and experience previously needed to leverage it.
The process couldn’t appear simpler:
1. Build a data warehouse – or don’t, your data is probably fine where it is.
2. Connect your data source to Power BI.
3. Build some visual dashboards; couple of clicks for each. Don’t like one, delete, it’s no biggie to replace.
4. Build some more dashboards, save them, share them, explore.
5. Use natural language to query the dashboards
6. There you have it. Big Data 101
This opens Big Data to the non-data science masses. Borrowing the strap-line from the Alaska Airlines chap ‘start small, think big’. With minimal investment organisations can dip their toe in the water (extend analogy to Data Lake as you see fit), realise some cheap quick win benefits and ROI almost immediately before extending their programme outwards.
Historically, when working with customers who have big data sets in a database somewhere and they know there’s value I’ve used SSRS in Azure to deliver the business information locked away in that data to influence strategy. Time-consuming and expensive. Cortana Analytics Suite is likely to make that step cheaper for me. Cheaper for me, cheaper for my customers.
Internet of Things
Some scary stats shown on screen – more IoT devices than people by 2020. Creating an estimated value of $3 Trillion.
Recently crudely described to me by an old school techie as “just Embedded-C for the internet generation”. IoT is already being used to change the world, and there is apparently no ceiling. As with all new technical ventures, it’s the setting up and getting started that takes the time and significantly eats into budget. Again, seeing someone show in a short presentation, in real time how Azure speeds up to a ridiculous extent your deployment of IoT is game changing.
It’s perfectly feasible to think 5 years ago if you had an amazing idea and happened to find cash in your pocket you could spend 8/12/16 weeks preparing your environment before you ever write one line of code. But now, have an idea over breakfast, spin up the right flavour of Azure for you, walk the dog and by the time you return Azure’s asking what took you so long.
It’s what it boils down to 99% of the time.
Microsoft has been so aggressive on reducing storage costs, which is awesome. Compute cost is the fun one with FDs. Yes, we can estimate costs based on 24/7 usage, and yes we can schedule, monitor and cap costs, but it has always been a fluid cost as it’s so easy to make changes to your environment that can affect costs.
MS announced yesterday the Azure Compute Pre-Purchase Plan. Phenomenal. We now have the flexibility to buy our compute costs up front, at potentially significantly less cost. Although this is unlikely to allow for 100% forecastable compute costs, at least we now we can buy and forecast a large proportion of our compute costs and then it’s the minority of costs that need managing at month end.
What this means
All of this is great for customers. But there are some hidden benefits for me too. I work a lot with organisations looking to change the way they manage their projects and use SharePoint (when done right, SharePoint delivers staggering PPM capability). We can show customers demos of our solution and talk to them about their problems in one meeting, but everything else requires follow up.
With a few minutes and a few clicks I can connect Power BI to the customer PPM data source at the start of the meeting, apply my pre-built Power BI content pack (dashboards), show my demo, then when that’s finished be able to talk to my customer there and then about raw, unfettered patterns we can see in their data.
Many enterprise customers I’ve met run projects on disparate spreadsheets – imagine connecting those together in a meeting, in five minutes, and exploring the findings with your customer together.
And this is just the start.