In this episode of the Hub & Spoken 100 episodes celebratory special series, Jason Foster looks back on conversations with business, data and technology leaders about deploying technology. We review how data platforms – and data products – enable delivery of value at scale, and the best way to go about that.
Listen to this episode on Spotify, iTunes, and Stitcher. You can also catch up on the previous episodes of the Hub & Spoken podcast when you subscribe.
What are your thoughts on data platforms and data products? We’d love to hear from you; join the #HubandSpoken discussion and let us know on Twitter and LinkedIn.
For more take a look at our on demand webinar: How to Build a Data Platform
Or you might like our whitepaper: Exploring Data Product Management
Accelerate returns from data at pace, and with agility, by using technology the right way. It’s one of the enablers to a successful data strategy, but can equally be an inhibitor if not implemented correctly.
[02:06] The need to build and nurture data products as a way of more quickly delivering value from technology.
[09:43] The emerging concept of data ops.
[14:00] What is AI and where are we with it?
[21:54] How much Tinder is building data products, how they’re doing it, how they’re managing the data platform and the approach they found that works best for them?
[27:35] The use of augmented analytics to support improvement in people’s lives.
DataOps is really an opportunity to bring some of the really big, important changes together under a single banner, and then give it a meaning and a focus. It’s nothing particularly new in itself, but it is actually a combination of three things:
(1) Automation. There’s still a level of automation that should be explored and used inside data, and data scientist and engineering.
(2) Agile agenda. Making change happen through agile practice, as opposed to traditional STLC and multiple approaches.
(3) Manufacturing. Getting straight through processing correctly, and your shortest possible distance from left to right.
Making things efficient in a nutshell, it’s these three things. That’s how you start to get change happening and support what you want to do.
There needs to be a maturing of the understanding of what AI is, and where we are with it at the moment. If you are adopting AI solutions at this point in time, you’re an early adopter and the majority of these projects will likely fail because of that. The reality is that the technology itself isn’t as mature as we might believe.
Over the next 10 to 20 years, organisations will mature as fast as technology will, and we’ll see some disruption; not only in low level work, but in higher functioning work – and in society.
Treating data as a product means bringing product management practices into data solutions. It’s around agile methodologies and treating data as a product – in a similar way to the regular product that you build.
At Tinder, the way they think of data as a product has evolved. The data teams are an enabler, and provide a platform to deliver data products. Though the consumer teams are then going to be owning these data products; they’ll be doing analysis and additional transformations on them. They own the data assets and they take it from there.
Using technology and analytics and data in AI should be for the benefit of freeing up time for workers. It’s one of the best things organisations can do. It doesn’t mean that AI will take over human’s jobs, but it’s the opposite.
As technology arises, people can focus more on their profession. People can be more of a consultant or can be more empathetic. They can work on their leadership or can focus on their customers. That’s how humans should be spending their time. People shouldn’t be spending their time on repetitive work or things that should be done by algorithms or machines. We should let the algorithm and the machine do what they do best and humans can do what they do best.
Technology can both be a boon and bane in building a data strategy. It solely depends on how we view and use it, whether it be in our businesses or personal lives. Regardless of all the advancement it brings, we should not let it take over us. Rather, it should be a tool that we use to have a better life and help us do what we do best and to be more human.
If you would like to review the full episodes of the podcasts included in this special episode, you can find them here:
Episode 57, Building global data products at HSBC, with Ranil Boteju
Episode 71, Understanding DataOps and its role in data strategy, with Rob Kellaway
Episode 34, Reality Check on AI, with Theo Priestley
Episode 87, How Tinder Approaches Treating Data As A Product, with Murali Bhogavalli
Episode 59, How technology and analytics can impact on human potential, with Gustavo Canton