Oracle puts effort into selling at HEUG, which makes sense because we are all users of Oracle software. It is interesting to know what is out there, but the likelihood is we will never get to use most of it. Also, Oracle pushes cloud which I find quite depressing because I don’t know where I fit in to the new cloud paradigm.
I made notes from a couple of these sessions, which are quite short, so I will combine them here.
Oracle Cloud Analytics
In my mind this is like OBIEE, but moved on a bit to be more user friendly, more drag and drop. It looks like Google Analytics, and is supposed to be just as friendly. The sales man promised a lot, that you could upload a spreadsheet, then it would automatically identify all the columns, and tell you about them, and tell you interesting information.
The data visualisation is very pretty, and it does seem user friendly. However, it strikes me that you still need to understand statistics, the underlying dataset(s) and the limitations of each to understand what it is telling you. The dataset used to demonstrate was a list of employees and contained a column called attrition. This doesn’t make sense to me because an employee is either working or has left. Attrition is something that pertains to a group over time. He demonstrated attrition against pay, which again doesn’t make sense to me because if an employee has left, they are no longer getting pay rises. My colleague pointed out that while the tool can find correlation between whether people have left and other data points, this doesn’t mean that the correlation found is actually the reason for it.
He demonstrated importing a spreadsheet and the tool detecting what the data in the columns is, and describing information about them. My suspicion is that this will miss the subtleties of the information in the spreadsheet, and end up being misleading.
It seems it can build reports from natural language queries, and also give natural language insights into data.
So while it is a pretty tool, and seems to reduce the technical knowledge required, it will still require statistics and knowledge of the underlying data to be able to use effectively.
Blockchain IoT and AI
The last session of the conference was stuffed with buzzwords to keep us all interested. It was led by Peter Szegedi - senior technology solution engineer - Oracle Digital EMEA.
Peter explained the life cycle of these products. They start off as a platform, I assume where they deliver libraries for to work with things, to applications where presumably it is ready to use in specific cases. It them moves to specific applications when it is mature. He didn’t explain what happens to customers who are using the platform services and don’t want to use the applications.
AI (Artificial Intelligence)
He considers AI is a moving target. Previously a chess playing program was AI, but this isn’t the case any more. Now we talk about machine and deep learning as AI. He Listed a load of algorithms that Oracle use. For some it is not a stand alone service but works across a number of applications. They all have cloud in their name, Oracle marketing is consistent on this at least! Though I noticed a number in the database. Some examples were Spark, R, Data Miner, some SQL functions.
He mentioned storing exam results as a use case for blockchain. This is something that was mentioned earlier in a keynote. He considered that this was an ideal use of blockchain. Later an insightful question mentioned that storing this information in one suppliers cloud controlled by one institution effectively undoes the advantage of blockchain - the fact that “everyone” agrees on the history. If Oracle Cloud controls most of the copies, it can rewrite history. Peter acknowledged this and said the blockchain would need to be spread across several institutions, but didn’t address the point that Oracle are selling this as a cloud solution. Another questioner asked about the environmental impact of blockchain, as it uses enormous amounts of energy, which again was acknowledged as an issue but not addressed.
Blockchain is a platform service which suggests Oracle haven’t quite worked out the best way use it yet.
IOT (Internet of Things)
Peter gave an example from manufacturing. Knowing which parts were installed in which cars allows a company to know which cars to recall due to a faulty part.
Another example is a haulage company tracking ships knowing when they will dock and having transport capacity to sell at the right time so they made loads of money.
To get full value of IoT, devices have to talk to each other and exchange information.
Passive IoT is e.g. a bar code, active IoT always connected, e.g. self driving cars communicating their intentions with each other. Students might not want to be tracked in this way. Maybe they could be encouraged to share the information?
IoT is no longer a platform service - it is in the applications.
At the end was the obligatory advert for Oracle Cloud. There is a 60 day trial. Some services are free after this period. You can play with the blockchain in the 60 day trial. This was also mentioned in the cloud analytics session. Apparently most people that sign up for the trial don’t make the time to play with the technologies before the period expires, which is understandable I suppose.