In this second instalment of the CoreBrief interview series, I talk to a pair of London-based technology professionals who have a passion for the pragmatic real-world applications of science and technology. Combined with their entrepreneurial spirit and industry insights, this interview presents a thought provoking exploration as well as an invitation to connect, collaborate and co-create.
Ali is an Entrepreneur and Tech Lead with 15 years of experience in building teams and products at some of the leading startups in the UK and elsewhere.
Automating a common challenge
JD: Ainy and Ali, I know you guys have been putting a lot of effort in your own time into an interesting project related to automatic generation of certain types of mobile apps. What can you share about it? How did that come about? What kinds of partnerships or other resources would be helpful to you at the current stage?
Ainy: About 175,000 people take the Life in the UK test every year to qualify for residence. I was one of them last year. I tried several apps on the App Store to help me prepare for the test but didn’t like any of them so myself and Ali decided to build one.
We did some market research and quickly realised that about 4 out of the top 20 best-selling apps on the App Store were for test-preparation. At this point, Ali took a step back and built an engine that transforms a spreadsheet containing multiple-choice questions into a beautiful, singing-and-dancing test-preparation app for the iPhone.
We are now dog-fooding the engine to build an app for the Life in the UK test – which was the plan all along. This will probably be followed by more test-preparation apps until we feel we have something reproducible. At this point, we plan to launch the engine as a no-code solution for building test-preparation apps.
Ali: We would love to partner with people who have: 1) domain knowledge about a particular little-known MCQ-style (multiple-choice) test taken by several thousand people annually and 2) the gumption to put content in a spreadsheet and benefit. If you are such a person please reach out.
Some examples of MCQ-style tests taken by thousands of people annually include: AWS certifications; regular driving license; driving license for lorries and buses; residence qualification tests (like Life in the UK); foreign language proficiency assessments like the TOEFL, etc.
On getting into tech and mastering new skills
JD: This is really awesome – I know you’ve already reached significant milestones with this project. I myself am taking the HSK (Chinese proficiency) tests later this year and could really benefit from a reliable and simple application that will allow me to practice with real questions – especially if it can be tailored to my current level of studies. I can imagine dozens of other context where this could be useful.
Ainy, what’s was the most challenging part of getting into technology and software engineering from another field? What helped you most on this journey?
Ainy: After a long hiatus from my professional career – digital marketing and project management, I felt a bit detached from my actual field and wanted to have a fresh start. Ali convinced me that learning front end development might be just the right thing for me. Perks like working on my own hours and creative aspects of developing the visible part of a website seemed exciting.
I started with an online course in HTML, CSS and JS. I liked how it started and I learnt quite a lot in a short time. But, as it went further, things started getting more complex. Understanding was one thing but when it came to the assignments, I found myself stuck.
I was about to give up when Ali pointed out that since I had been learning remotely, I wasn’t able to see other students struggle – which is important to overcome imposter syndrome.
I enrolled in a real-world bootcamp. This was much better. Interacting with other students showed me that the struggle was mutual. Very much like school, when there is one student in every class who understands every lecture, I sometimes found myself to be that student. This led me to helping other students which was a big confidence-boost that reinforced my self-belief and encouraged me to go on.
Why can’t you subscribe to the content you care about on your Kindle?
JD: What other interesting side-projects or ideas have occupied your attention recently?
Ali: 15 years ago, if you wanted to share your ideas you would write a book or a blog. Now you can write a book, a blog, a tweet-storm or a newsletter, record a podcast, a live-stream or a webinar etc.
I have been thinking deeply about the explosion in the number of viable channels for distributing value over the internet and how to use this as leverage. For example, there are ~100M people who use Kindle to primarily read books. Why is there no “read-it-later” tool that delivers beautifully typed, Kindle optimised articles to your device at the touch of a button?
Ainy: I am planning to restructure my writing into a form of a blog. Over a period of time I have written some content that has been lying around. I think now is a good time to work on it and maybe get it published online.
A good time to innovate and leverage opportunities
JD: It sure is a good time to take a step back and focus on things that are important to us, or that we’ve been meaning to do for a long time. From the point of view of technology, the workplace, business and society, what do you see as some short-term and long-term consequences of the current pandemic, the lock-down and the expected global recession? Do you see any emerging winners from the whole situation?
Ali: The most useful thing to come out of the pandemic for me is learning that hyper-efficiency is fragile. This applies to people, business and society and so belongs to a category of rare ideas that are scale invariant.
If I were not cynical I would bet that people, businesses and society would learn this as well and correct course e.g. to support more local businesses instead of long, thin, “Around the world in Eighty days” supply chains. But I am cynical and so I’d reckon – in short-to-medium term – COVID-19 would likely tip the scales further in favour of big corporations leading to a further increase in hyper-efficiency and fragility.
That said, long-term is where things get more interesting because I cannot help but bet on makers to innovate and leverage opportunities presented by the “new normal” before big corporations could decide on which video conferencing tool they would like to use for stand-ups.
For example, online learning where innovation up until now has meant replacing the teacher with a talking head and the ambient interactions that happen in the classroom with a Slack channel. This just about works for e.g. front-end devs looking to ramp up on a flavour-of-the-month framework. But it is sub-optimal for e.g. primary school students who are now having to rely on online learning as the only recourse to education.
JD: In your view, what technology or software trends hold a lot of potential but are particularly under-appreciated / under-reported areas worth exploring further?
Ali: I have always maintained that the technology opinion-sphere is a FAANG (“Big Tech“) monopoly. A large chunk of the value my employers and clients have derived out of my work has been me telling them: “Look, you’ve only got a handful of paying customers?” – usually in response to statements such as: “We are moving to AWS because we need to scale”, “We are using Spark so that we can run out ETL on a petabyte of data”, “We need to use React because Facebook uses it” etc.
There is a rant/blog in me about “What can be done with a $5 Linode and a $10 PostgreSQL instance?” I suspect the answer might surprise quite a few people.
In one of his talks, Nassim Taleb describes what is called the Lindy effect as: “The life expectancy of a play running on Broadway is proportional to the time for which it has been running already”. In terms of software, if a product, tool or service has been around for 20 years, there must be a good reason for it and it is likely that it will still be around for another 20 years.
So my bet is on classical, proven technologies like Unix, bash, vim, C, PostgreSQL, Redis, Python, the file system, the network etc. Or fundamentals like: How to write well? What does a good pull request looks like? What does a good commit look like? What does a good code review look like?
Is Data Science just another bubble trend?
JD: Ali, you’ve worked on both software engineering and data science projects. From my experience, while these fields are increasingly closely related, there is still some disconnect. What are your views on this subject?
Ali: Anecdotally, I can state that 90% of the businesses that have in-house Data Science teams are incapable of deriving any business value from Data Science outcomes. This is because of two main reasons: 1) they do not have the data to enable good Data Science outcomes, and 2) they do not actually have a market to sell Data Science outcomes.
Also anecdotally: of the remaining 10% businesses, 90% can replace the entirety of their Data Science with a few nifty SQL queries that involve judicious use of `regr_*` functions and I’m confident their customers won’t even know the difference.
After having lobbed that grenade in the mix, I am ready to talk about the remaining 1% of businesses who 1) have data to enable good Data Science outcomes, 2) have a market to sell Data Science outcomes, and 3) the Data Science outcomes require something a little more involved than SQL.
My hypothesis is that 90% of these remaining businesses can achieve their business outcomes by asking their engineers to do a slightly intelligent version of what tpot does. If people can write APIs without knowing how Diffie-Hellman key exchange works, they can also put a basic recommendation engine in production without mastering the underlying mathematics.
With that said, there is still room for Data Science specialists not only for the remaining 0.1% businesses but also for writing SciPy modules, for example, to encapsulate the latest and greatest research in ML. There is value in Data Science specialists doing actual Data Science instead of pseudo-engineering.
The key to switching for a career in IT to entrepreneurship
JD: What does it take to go from a good software engineer or IT professional to a technology entrepreneur?
Ali: Simple: 1) Stop trying to impress your peers, and 2) Start trying to impress your customers.
On a side note, I believe moving from 1) to 2) is good for career advancement in the general case. Google “Expert problem” to find out more about this.
Who to follow on Twitter
JD: I know you’re active on Twitter (@alixedi). What are some higher quality Twitter accounts you’ve found consistently worthwhile following?
Ali: I have been really enjoying @dvassallo lately. He does experiments ranging from marketing campaigns, shorting the stock market and transparently shares his outcomes.
I would love to do the same one day – when I have something nearly as interesting to share.
How to use your calendar to reach new levels of structure and productivity in your life
JD: Do you have any productivity tips that you’ve recently found practical and useful?
Ali: I have been a productivity nerd since as far as I can remember but all the tools, techniques and hacks that I have used so far have an underlying assumption that productivity is an outcome instead of a signal.
You want to get more done? Put all the things that you need to do in a To-do list and let your fight-or-flight response push you into staying up until 2am to get it all done.
I have come to strongly disagree with this notion – to an extent that I think its irresponsible to let people mess with their minds like this.
Perhaps less importantly in this case, it also doesn’t work. Depending on the number of days since my last holidays, I lasted anywhere between a couple of weeks to a month before burning out.
Lately, I have started experimenting with a new system:
1) I put slots for all the things that need doing in a calendar e.g. run, email, daily stand-up, retrospectives, planning, writing etc.
2) I try and do what the calendar says.
3) Because life, sometimes I can’t so at the end of the day, I do a review in which I retrospectively change the calendar for the day to reflect what happened e.g. If I didn’t go for a run because I got sucked into checking my email, I delete “run” from my calendar and put “email” instead.
This means that every future event in the calendar represents a plan and every past event represents a log.
At the end of the month, I parse the .ical file and produce a diff between my plans and my logs. The diff is a signal e.g. if it is greater than say 20%, I revise my plans.
Personally, this has been life changing in that I am productive without paying for it with my mental health.
For more actionable thought leadership by Ali, Ainy and others, subscribe to this blog. Comment or leave feedback to engage in discussion or for support and encouragement.
If you know someone else you’d like to see interviewed here, do get in touch.