With predictability and the ‘need for speed’ presenting major challenges, Mark Porter explains why R&D in tech can no longer be a black box.
Mark Porter is the CTO of database software company MongoDB. A developer at heart, Porter has spent the best part of four decades working around databases on the vendor side with AWS and Oracle, but also on the customer side at NASA, NewsCorp and Grab.
At MongoDB, Porter’s role as CTO sees him lead the engineering organisation and he is responsible for the long-term technical vision for the company.
“What that looks like on a day-to-day basis includes many things,” he told SiliconRepublic.com. “On the technical side I drive our three to five-year technical vision, identifying and addressing gaps in our offerings.
“I also spend a big chunk of my time talking to C-suite executives worldwide to communicate this long-term vision and understand their needs.”
‘One of the biggest challenges we see with digital transformation is what I call the innovation tax’
– MARK PORTER
What are some of the biggest challenges you’re facing in the current IT landscape?
Innovation at the speed needed by the market is the biggest challenge. We are past the point of debating an organisation’s dependence on software, but what is now up for discussion is how an organisation makes innovation a key differentiator and how they build and scale modern applications.
The speed at which a new application can be deployed, along with the number of innovative features it includes, has a direct correlation to business success. The days of outsourcing your most innovative projects are over. If you do, those folks basically have to function as empowered and engaged employees of your own company, which is challenging.
Surprisingly to many, I’ll rank predictability of engineering as the second challenge. As software has become a more central part of almost every business, no matter what the industry, it has become more important for this part of the company to become more predictable – not something that software engineering groups are known for, frankly.
Predictability of cost and results is expected of every business unit in a corporation today, or it can’t move fast and be competitive. It’s no longer OK for R&D to be a black box and say, ‘the new release will be done when it’s done’. If your engineering team is still living in that world, you’re leaving a lot of opportunity on the floor.
So, how do you do that? You need to incorporate key software techniques like small teams, microservices, local testing, etc. In other words, everything you’ve heard in the ‘shift left’ movement is actually about producing more predictably high-quality software on the agreed-upon schedule – or at least knowing it’s off the rails as soon as possible.
What are your thoughts on digital transformation?
One of the biggest challenges we see with digital transformation is what I call ‘the innovation tax’.
This metaphor works better than you might think. Like a tax, you have to pay it, and like taxes that go unpaid, the longer you don’t pay it down, the more it expands from both interest and penalties.
The innovation tax is the slowdown in velocity that occurs from a combination of using old technologies, blindly using technologies not fit for purpose and, frankly, just having too many parts in the technology machine that runs your company.
It can occur at the micro level, where a team has too many frameworks or APIs, or at the macro level, where departments depend on internal details of the products produced by other departments. So, things just get slow. And it’s not clear why, but everybody feels it. The corporate mud which used to be around your knees is quickly approaching your nostrils.
There is a tipping point for a team when the tax has become too high for either the developers to pay or the management to understand. And this is why many projects have insane schedule slips, outright fail, or even declare bankruptcy and start over, sadly doomed to repeat their mistakes because leadership rarely understands that it was them that failed, not the employees.
The better way is for organisations to stay conscious of all their debts – technical debt, architectural debt, organisational debt, Peter’s Principle debt, and all the others. Software engineering teams could learn a lot from the banking industry and how portfolios of investments at different stages of maturity have different needs.
What big tech trends do you believe are changing the world?
20 years ago, the world was full of huge tube TVs, and they hadn’t changed in meaningful ways for many years. Then, out of the blue, flat-screen TVs revolutionised them, and the innovation is still going.
Databases around the turn of the millennium were the same way. Then modern databases arrived and, like flat-screen TVs, the innovation just won’t stop. We have modern document databases that handle data types at speeds and sizes that were previously unfathomable, and that are easier to program against than anybody ever knew they could be.
They scale up and scale down as needed, and instead of being tied to one platform, run everywhere. One of the reasons I’ve been fascinated by databases for more than 30 years is that the promises they make in terms of durability, correctness and availability are some of the hardest promises to make in computer science – and now we’re combining those with massive speed and scale.
How can we address the security challenges currently facing your industry?
Security, while always important, has become job one for any company that deals in personal information or financial data. One of the basic building blocks of that security is encryption – scrambling data with a secret key so that only authorised users can read it.
But when you encrypt your data, you lose the ability to do searches and queries on it, which makes writing secure and performant applications slow and error prone. Or you can choose to encrypt it over the wire and on disk, but leave it unencrypted in memory, which means that your cloud provider, their DBAs, or anybody who has exploited them, can read your data at will. What are CTOs and CISOs to do?
Enter queryable encryption. With this new technology, which we’ve been working on for years and just released the first version of, you can store your data and query it at speed, and yet be able to prove that nobody, not even your cloud provider or their operations personnel, can read it.
By doing this, we’re removing what used to be a tough choice – whether to protect your data fully or use it efficiently – and let people just write secure and performant apps.
10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.