The CTO of iFlix talks about the first steps a startup should take when looking at cloud: Play with it, try and scr*w it up, tear it down and start again
We know cloud computing and the startup world is a match made in heaven. In spite of it, startups do not know how to make the most of the cloud ecosystem or how to use it to scale fast or even the first steps to take when exploring the cloud.
Here are the edited excerpts:
What has been your cloud journey? What are the challenges it has helped you overcome?
We started iFlix in September last year. I was the third employee and, today, we are close to 150. We wanted to launch our product within six months of starting, and that is a massive challenge because we are talking of a product that could potentially serve millions of people.
So it should be able to scale exponentially and provide quality services like pulling studio content and delivering the content to customers in a reliable way, dealing with bad bandwidth, among other things. To be able to do that without even a team to begin with is a big challenge.
To build something that will deliver TV off the Internet is a pretty new business model and I cannot think of a way it could exist without partnering with Amazon Web Services (AWS) or any other capable cloud provider.
Just the capital expenditure required to build the hardware that will support peak traffic, but will be unused 95 per cent of the time at its full, is huge. It will just not be feasible without cloud.
We provide our demand services to customers, about 11,000 hours of content, for US$2.5 per hour, which is a fraction of what the big players do at the moment. To be able to do so, we leverage a lot on AWS and we are going to go deeper.
With cloud, we have been able to build with a small team what would traditionally require a large setup.
How big is your technology team?
About 30 right now, and I really would not want that to go too far beyond. Let’s say a year’s time from now, I would like to have a ceiling of 50 people. Once you have more people than that in a team and you want an end-to-end view, you start becoming a people management company with a big overhead.
I think, with automation, open source solutions and cloud capabilities, you can keep the smallest team possible to deliver the biggest outcome.
Headcount is not a KPI, contrary to what most companies think.
What is the biggest mistake you have made in your cloud adoption?
We have not been around long enough. We have had a blank sheet of paper to set up what we have now. Usually, mistakes come from transformation, but we had no legacy. Anything that has not been quite right… we could have done slightly better. But there has been no catastrophe.
We (the tech leadership) come from a strong background of scaling up and we have learned a few things the hard way in our journey, so the biggies are sorted out.
I am not saying it is perfect. It is not perfect in any shape or form. We may make a mistake going forward. However, we have been largely successful so far.
In your interaction with startups, what is the one thing they are doing wrong in embracing cloud?
Companies just try to do too much. Cloud is an area they should explore if they want to put in minimum effort for maximum output. This is because a lot of things come pre-packaged… [The] work is already done for you.
A lot of startups would typically focus on getting machines… Infrastructure-as-a-Service kind of stuff, and then do everything else by themselves. They will get an admin guy who understands the operating system or a database guy who will then build on top of the box provided by a cloud partner.
The one thing I would recommend startups to do is ensure your senior team is strong. Employ senior generalists with diverse backgrounds and get them to look at services out of the box that AWS provides — the reason I say AWS is because others don’t do it anywhere close to the degree that AWS does – to solve a really hard problem.
Just on an operation level, you can overcome massive challenges for a fraction of the cost. On the software level, you can pretty much do the development today and have it in production tomorrow. That’s a massive competitive advantage.
So, getting people who really understand software architecture in the context of hardware architecture to review all the services available to them via cloud to see how they fit is crucial. Get AWS to do to your heavy lifting. Most companies do not leverage it close to what they should.
For a startup just starting to look at cloud what are the first few steps?
- Get your hands on the system
- Play with it
- Try and scr*w it up
- Tear it down and start again
- Don’t add multiple layers. Just get one box and explore its potential
Get something to work on that box end-to-end and use it to get your product out as fast as possible. It does not need to be perfect. There is a famous quote that says, ‘If you are not embarrassed by your first release, you have released it too late’. Use the cloud to get it out there quickly
- Lastly, continue to iterate fast.
The post Get your hands on the cloud and try to scr*w it up: Ash Crick of iFlix appeared first on e27.
from e27 http://ift.tt/1YkbjjE