The Studio Blog

Weekly Journal: Week 3, 2019

Last week was fun in a brain-melting kind of way…

Data Science and Neural Networks

I found some great reading on using PyTorch to build neural network classifiers and also a bunch of great writing on using Word2Vec – the problem I was trying to solve was how best to design the inputs and outputs of a NN classifier. There are so many potential data-points that could be used to influence classification, where to start?

I ended up choosing a simple, yet not-quite-so-naive approach of usingTF-IDFto build up a vector that can represent word-embedding after normalising the text-vocabulary withSpaCy. Then we pass in product information to get a predicted product classification (category-confidence vector) …it took a lot of ruminating and reading to get to this conclusion. This week I aim to write the data transformer to build the training set from the data available.

I spent a lot of time last week tuning my fuzzy-match engine as well. As I mentioned last week, an article is in progress. This week I have made the code multi-processor so at this point it is ready for a write up…


I finished reading theWolfenstein 3D Black Book. This book is basically a code-review of one of the most influential games in history. It was riveting. I could not put it down and basically finished it in a week. Loved every second. There is something really compelling about discovering how a true genius of our time (John Carmack) thinks and solved his programming problems. This is the same reason I used to lovedisassembling demosand games as a kid to understand how visual effects were made.


Just sharing one of my goals this year is “Fitness every day with a focus on diet”…I will share more goals as it makes sense to. Friday was a 14 hour day at work and the first day I missed doing exercise. I missed not doing exercise after three weeks of sticking to this goal. Got back in the saddle on Saturday and Sunday…

Weekly Journal: Week 2, 2019

Last week was interesting – I had a bunch of things going on at once and all of them were kind of interesting. In fact, it would not be hard to write a full essay on each of these.

Weekly Journal: Week 1, 2019

I decided I would share weekly insights with you as a way to reflect on the week that was. I felt like there is so much we do every week that could be discussed, shared, considered further. I wanted a place for that to happen and so this is it…


Getting Data in the Enterprise

So many organisations today struggle with their data. The world of big data, analytics and predictive modelling is so attractive, yet, the perceived complexity of achieving a data nirvana is so great that data projects either spiral out of control or never start. So how can you get a handle on your current data, whip it into shape and start using it?


Managing Complexity

No matter what industry you’re in, you have to deal with complexity on a daily basis. From the complexity of computer systems you use to get work done, to the complexity of the projects you’re managing – it seems a byproduct of the information age is complexity.


Delivering Results: Software Estimation

More often than not when engaging a software developer there will be a conversation about price:

“Here is what I want. How much will it cost?”

The conversation may be more subtle, or nuanced, but the bottom line is pretty much the same no matter what: A true definition of what is actually being asked for is not very well defined at the point in time a price is being requested.

Where can you go from here?


The Future of Corporate Training

I started working on different approaches to corporate education and training with Thinking.School in 2016. Initially, I started developing a program that was quite abstract: It was set up around ‘thinking strategies’ and how they could be used in problem-solving. The problem with abstract training like this is that we end up with a metaphoric hammer looking for a nail. It was a set of strategies looking for problems to deal with. More interesting, I found is that this problem with abstraction in training is not unique. As I spent more time working in the space and talking to others, I saw the same issue plague enterprise scale Design Thinking programs – they make sense, they sound great, people enjoy them – while they have pockets of influence, they struggle to have a broad lasting impact.


Designing for Design Patterns

As we worked with financial services and insurance companies over the last five years, we saw that these knowledge-gaps at the convergence of design, development and security were something that we could solve with Design Patterns. So we started doing the research.


Modernising Legacy Applications

There are a bunch of different kinds of ‘legacy’ systems which require different strategies to decouple such that an ‘API’ can be provided for a new experience to be built over the top. In this post we are going to look at a set of common legacy system design patterns and how you can tackle modernising them.

Level 25, 88 Phillip Street
NSW 2000

+61 2 8024 5975
[email protected]

©2003-2019 Pixolut Pty Ltd trading as Thinking.Studio
All Rights Reserved