Technology roadmap, 2019–2020.
v0.1: We’re building some (surprisingly sophisticated) video education technology to support our new work across Southeast Asia this year.
Here’s a Spotify playlist to make this big ol’ post more fun.
This post contains:
- The basics of the Fitzroy Academy;
- We are doing a ‘capital G Good, capital T Thing’;
- We are making an online education platform;
- Delivering tech across multiple countries;
- Data collection, privacy and ethics;
- Important questions we want to answer;
- We’re open sourcing almost everything;
- Distributed, remote teams and work structure;
- Engineering detail (aka. the Github readme);
- Flipped learning: What it means and how it works;
- Aligning with the UN Sustainable Development Goals;
- Hiring: Middle-to-heavyweight Python dev
- Hiring: A data scientist contractor
- What have we missed?
The basics of the Fitzroy Academy:
“We make online learning for social entrepreneurs, teachers and the ecosystem. We succeed when our students and teachers love our resources, and the ecosystem can measure our impact.”
Our core values centre on broadly accessible, people-first education and tech.
In the last 4 years we’ve built ~30 high quality video lessons on entrepreneurship, social impact and digital skills.
We are doing a ‘capital G good, capital T thing’;
Reid (our devops guy) thought it was important to point out that we are actually-and-definitely-truly a social impact organisation.
We’re backed by multiple high impact philanthropic funds, are wary of VC capital and will always put people before profit. We’re in it for the long haul.
Jackie, GM @ Telstra Foundation says it like this:
“They’re a brilliant team and they have made something amazing which is exactly what the sector needs.”
Our work helps women in developing countries earn more money, results in fewer small businesses failing, better educated kids, and happier teachers. 😍
We are making an online education platform:
Heres are few screen grabs of prototypes from early April 2019:
The exact architecture and deliverables of this platform are constantly changing, and figured out in partnership with colleagues and friends.
Q: Aren’t there a bunch of Learning Management Systems, open source and otherwise, out there already? Why re-invent the wheel?
A: Yes, platforms like Litmos and Moodle already exist.
However, we have strange needs that make existing platforms unsuitable.
For example, we need:
- To deliver learning at ‘close to free’ for less materially rich communities.
- Seamless translation across multiple languages within the app, overdubs, subtitles, and even within segments of lessons.
- Incredibly strong data security and user-led data control, especially considering the GDPR and <16 year old students.
- Fine grained content controls to allow us to work within countries with strict policies, including oppressive regimes with high levels of corruption.
- Flipped learning stuff for under-resourced teachers.
Delivering tech across multiple countries:
One particular quirk of our work in Southeast Asia is translation and internationalisation. This means language, context, and politics.
We will be delivering lessons filmed in Khmer that need to be translated into Tagalog, English, and other languages and then delivered in other contexts.
Our partners work in places with wildly varying levels of internet penetration, education standards, and cultural norms. This makes the tech tricky!
The cross-cultural nature of our work demands that we have a distributed, diverse team of people working on the problem from every angle.
Data collection, privacy and ethics:
One beautiful thing about a new build is the chance to Do It Right™.
We’re planning to go above and beyond with data and privacy:
- Full GDPR compliance.
- Quality encryption and data storage.
- Transparency on what we collect and how we use it.
- Near-complete end user control, with download and delete permissions.
The last point is tricky. We want students to own and control their data, but we also need to collect in-depth data for teachers, institutes and researchers.
Our first core value puts it reasonably well:
“Right > rich”
Always choose ethics over money, then effectiveness over rightness.
The optimal balance of ‘effective’ and ‘right’ is an ongoing discussion. 🖖
Important questions we want to answer.
As Caroline Fiennes says in her evidence based interventions lesson:
“Ask an important question and answer it reliably.”
We want the data we collect to help define and answer some questions that are useful for the wider ecosystem, outside of our own efficiency and operational concerns:
- What learning is most useful, and why, and when, and where, for whom?
- What elements outside the online learning gives the best bang for buck, i.e. class size, program length, financial support, etc?
- What geographies or demographics are best for deployment, i.e. do we get better results in some places than others, and if so why?
- Etc.
The full list of Important Questions Worth Answering™ is still being built.
The big goal here is to help answer questions that everyone in our sector can use to do better work. We can’t do this alone, but we can play nicely with other researchers, geeks and institutes.
We’re open sourcing almost* everything.
Because:
- Speed.
- Accessibility.
- It’s more right.
Speed.
We don’t have time to mess about, and we don’t want to hide our IP behind lock and key where it can get stale. Too many people need support and an open source approach is a faster way to implement on and offline.
Accessibility.
As we’re trying to be effective across many contexts, we need to use simple, commonly understood tech to allow more people to be involved. This is why we’re not using any bleeding-edge code hipsterisms.
More right-ness.
Devs like open source, giving away good work is good geeky karma, and we want to work with good people, all around the world!
Good values beget good people, and we can’t do this work alone. 🥰
Note: We say ‘almost’, as there are some things that are dangerous to open source, so we’ll go as far as we can while being responsible.
Distributed, remote teams and work structure.
Our tiny little team currently includes:
- Board folks: Italy, Singapore, France, the USA, Cambodia, and Australia.
- Paid staff: Australia, Cambodia, and the USA.
- Partners, teachers and customers in many more places besides!
Therefore we have to work remote by default. If you’re a dev and would like to work with us, you must prefer to work remote.
Engineering detail (aka. the Github readme).
We’ve spent time thinking through our philosophy and approach to actually shipping code, and the important stuff is covered in our Github README.
While the README is built for devs, it should make sense to non-technical people, as long as you skip all the boring SDLC / service architecture stuff.
Flipped learning: What it means and how it works.
Flipped learning is a style of teaching built upon:
- One-to-many learning (lectures, videos, etc) is done at home or online.
- Many-to-many work is done together, in classes.
The simplest version of this is a lecturer providing a video of her lecture that students can watch in their own time, which frees up class time to do the practical coursework together with support from the rest of the class.
The term ‘flipped’ comes from flipping the work order. ‘Homework’ is effectively done in class, while lectures are watched at home.
We have a much longer post about how we do this, and we’ll be releasing more resources for teachers to do better flipped work in 2019–2020.
From a technical point of view, all our lessons are built around flipped methodology, and the ‘back and forth’ of good flipped forms a baseline for how we measure and report on effectiveness and learning outcomes.
For example, a Facebook ‘active user’ is someone who’s visiting the site every day, possibly every hour. For Netflix, binges are a good thing.
For education, bingeing is a dark pattern to be avoided, for the same reasons that ‘cramming’ before exams leads to low long term retention.
Put simply: The rules for education are different to those of advertising, media consumption, or news. That means the tech is different!
Aligning with the Sustainable Development Goals.
We’ll go into more detail on this elsewhere, suffice to say that the SDGs are a hecking good idea and factor into our work. Specifically:
#4: Quality education. (This is obviously the big one.)
#5: Gender equality.
#8: Decent work and economic growth.
#9: Industry, innovation and infrastructure.
Within the SDG idea of ‘sustainability’, we look the dimensions of human, intellectual, financial and social capital from an educator’s point of view.
That is: Learning, money, and impact; social enterprise in a nutshell!
Hiring: Middle-weight Python dev.
We’re looking for a remote, capable and kind developer. We’ve chosen Python as the primary back-end language because it’s widely known and lets us work with more people across the world.
It’s also one of the natural languages for data science, which is a focus for a significant portion of our technical roadmap. 🤙
Must-haves:
- Is a nice person;
- Python skillz;
- Has written API driven services before;
- Has written complete apps (we are starting from scratch);
- Has examples of things built from scratch we can look at;
- 2+ (or more?) straight years of developing API-driven services in the wild;
- Exposure / experience with serverless + microservices work;
- Good communication skills and remote working experience.
Nice-to-haves:
- Open source experience (can we see your github please?)
- Some comfort with front end and package management.
What the work looks like:
- Our entire team works ‘part time’, anywhere from 1/2 to 4 days a week.
- We have multiple vid-catchups per week for sprint planning and whatnot.
- We expect everyone in our team to be smart, self-contained, and have great communication + time management skills.
How to apply:
- Read this whole blog post
- Read our Github
- Read our code of conduct
- Email will@fitzroyacademy.com with links to your existing work
😎
Hiring: Data scientist advisor.
We need advice and support from someone who understands data.
This is very much a short term, contract pay-you-for-your-advice sort of deal.
We just need someone to sanity check that the data we are collecting will be useful, that we’re collecting it in a sensible format, and that we’ll be able to do useful things with it in the future.
How to apply:
- Read this whole blog post
- Read our Github
- Read our code of conduct
- Email will@fitzroyacademy.com and say hi!
🤓
What have we missed?
This is a v0.1 of this post, please email me directly via will@fitzroyacademy.com if we missed stuff.
