It’s interesting that the ancient Greeks understood that the revelations of the Oracles were not seen as the objective truth and they were fully aware of the unknowability of the divine. I think everyone went to the Oracle Analytics Partner Forum in Athens with some trepidation about what the future holds for Analytics & the Cloud @ Oracle Corporation. The news is good….
There were a number of important takeaways for partners & customers alike. I mostly focused on the “tech” stream – the sales stream was equally interesting, but I’ve always found in the world of consulting being able to speak firsthand about your experience in helping customers solve problems is more relevant in building rapport & trust – in other words, it’s important I know what I’m talking about.
Anyone who’s read my previous posts knows that I am quite a big fan of Oracle Analytics Cloud. At Qubix, our team have poked and prodded and tested it right up to the extreme edge cases – and we know that the platform works really well – it’s fast, it’s stable, it’s easy to configure and it’s easy to use. But, I’ve shared some concerns with our team that the desire to increase “self service” is a double edged sword.
We all “self service” driving our own cars – but imagine trying to drive that car without having any lessons first. Or, (speaking in a personal capacity), how bad your driving was in the first year of driving – you just don’t have the experience to know what corners you can cut, versus those that you absolutely must not. And the same applies here – self service is fantastic, but it needs to be within the right framework and with the right guidance.
So the good news is the the semantic layer of the Oracle Analytics Cloud is here to stay, but that Data Visualisation Cloud Services continues to go from strength to strength. Some of the new capabilities and features of DVCS are nothing short of awesome.
Machine Learning & Artificial Intelligence
Mats Stellwall did a great presentation of what’s coming in the Machine Learning and AI spaces. I always love Mats’ presentations – his passion for solving complex problems and then sharing how he did it in an easy to understand and accessible fashion is always a pleasure. There’s never any games from him, he just knows his stuff really well and enjoys sharing it! This time, he was showing us the new ML and AI features.
One word – WOW! ML & AI has been around for years, but there’s been 2 blockers to it’s widestream adoption:
- Algorithm Complexity
- Computing Power
The maths isn’t “that” hard, but it’s always been quite inaccessible – so the new platform will be really powerful and make it possible for people to spend A LOT more time doing analysis and innovation, with a lot less time spent writing code.
Increasing integration between apps and tech (we’ve run PBCS => DVCS already and it’s good) means that there’ll be less time needing to be spent on the plumbing, and more time spent modelling the business and asking interesting questions.
Turning a Corner
It was a bit of a shock (not a surprise, a shock) when Oracle fell off the Gartner Magic Quadrant – a lot of people were wondering what the hell was going on. But let’s also be honest, OBIEE had it’s day. The semantic layer was and still is awesome, but you’ve got to implement it the right way. And Answers was really powerful, but the cost of change and effort to build dashboards that sometimes still lacked sizzle, was pretty big.
So it’s good to see the strategy to reinvent playing out now.
The OAC platform goes from strength to strength – the extensibility of the visualization library means if you can imagine it, you can build it. When you combine that with the horsepower offered by Essbase (see my previous blog), the scalability offered by the underlying RDBMS, it’s a pretty unbeatable platform. Combine that the some of the ML & AI capabilities just round the corner and it’s a world beating answer.
The next piece we took a look at was Autonomous Datawarehouse Cloud. This was a pleasant surprise – DBCS running 12C is already pretty smart, set it up right and it just runs and runs. But the real win was the step up in self tuning.
Autonomous data warehouse self tunes the database, and the example use cases were excellent.
Looking forward to getting my hands on that and seeing how far I can push it…tuning databases has always been something we’ve excelled at, so I’ll be interested to see if I can beat it.
This was actually a question someone asked me while we were there. Why Oracle at all? Their observation was that some of the Oracle offerings, especially on the big data, ML & AI side were just repackaging opensource products. Why not just go 100% open source?
I was pretty shocked by this question – the answers are pretty self evident:
- Single Stop Shop
- Infrastructure designed for the job
- Ultimately & Quickly scalable from 0 to as big as you can imaging
- Guaranteed to be compatible & “just work”
Then there’s more than just the open source products – DVCS, DBCS, Essbase, Big Data SQL Adaptor (which BTW is fantastic), ODI are just a few of the compelling answers to that question.