Building the Flipkart Flyte MP3 app, the 'Lean Startup' way
Learn from the experience of a team building a application for the indian market. And an agile iterative development process we follow, in our quest to get get millions of users :-)
This is not a deep technical talk, but will cover, in breadth, many aspects around engineering and releasing an app like this.
Flipkart launched Flyte, India’s first legal music download service. With Flyte, we were attempting to make a fairly large (and potentially painful) behavioural change – moving people from downloading pirated music to purchasing legal music. When tackling something new like this, we knew there would be a lot of experiments and learning.
I’ll try to cover:
History of the Flyte MP3 android application, and the goals behind it
Moving from a traditional release cycle to an agile, ‘lean startup’, model, with smaller, more frequent updates (~bi-weekly)
The metrics-based decision making process around product features
Some context on UX/UI and the design process we now follow
Tools and techniques that we use (and can recommend)
In general, things we learnt, things we wished we’d known, and things we are still figuring out.
Currently: I run the team at flipkart that builds mobile apps for digital content (the flyte brand). When i’m not running around being a pointy-haired boss, i help define some product features, write some code, and argue about the right way to use github.
In general: A technology enthusiast, having worked on many diverse technologies (Networking, Multimedia, Virtualization, Mobile, Location) at companies like VMware, Microsoft, and now at Flipkart. Prior to flipkart, I ran a startup that provided road traffic info for indian cities via an android app.