Requests per day
Time spent serving 90% of the feeds
Engagement increase using Tinybird
daily.dev is the fastest growing online community for developers to stay updated on the best developer news. Aggregated developer news from hundreds of sources every day is served to over 100,000 developers each week with a personalized feed based on each user’s preferences and the topics they choose to follow.
daily.dev collect and rank articles from hundreds of unique publications. Devs can bookmark posts, sync their data across devices, and read later whenever they want. The daily.dev web extension is currently available for Google Chrome, Microsoft Edge and Firefox. There's also a progressive web app (PWA) for mobile devices.
Developers using daily.dev generate around 150 million events each month as they read, upvote and discuss the latest news. daily.dev wanted the personalized feed generation process based on each developer’s preferences and behaviour to be super fast. They need to find the relevant inventory for the user, rank the stories and serve the personalized feed in as few milliseconds as possible.
daily.dev is growing fast and their existing process was struggling to scale with the number of events and requests. The personalised feed needs to be generated in real time, with the flexibility to try out different ideas, and that’s where Tinybird really made a difference.
The critical factor in switching to Tinybird was the developer experience for the people at daily.dev. Now they can simply write SQL and apply the processing both retroactively and in real time. They have the agility they need to scale their business. It’s only a matter of minutes between having an idea for A/B testing and putting it into production, without having to worry about the infrastructure or processing times. Working closely with the Tinybird team, daily.dev was able to optimise their data processing to aggregate and materialize the latest data as quickly and efficiently as possible.
Thanks to this ease-of-use they can now optimise the processing and personalization of the feed for every user. SQL logic in a single pipe (a series of queries) ranks the news article for the user to control which articles are pulled and served to the developer.
An optimally personalised feed for each developer led to a 60% increase in user engagement.