When AWS Lambda was introduced back in 2014, it was regarded as a cool tool to control and programmatically scale EC2 instances depending on traffic changes, process new files on S3 buckets and run simple automations when something happened in AWS in general. And then, out of the sudden, AWS API Gateway came into play and changed the way we think of serverless.
The rise of the serverless backends has begun. With the WebSocket API, competition from CDN providers (like CloudFlare), specialized FaaS (like Twilio Functions) we can be considering this “a landscape” of its own.
This said we can’t not hear the voices of companies that invested heavily into serverless with their API’s and built mobile/webapp backends directly in this model. Yes, there are major success stories, but these aside you see more and more situations where complexity ruins the whole idea of serverless infrastructure. Stories of thousands of dollars spent because of a small error in the code. Full-platform hiccups due to overlapping cache misses. Couple thousand lines of stack configurations just pass the data to another function!
Yet again it seems that you cannot really build a general use platform on top of a specific use. That beast of a system will be lurking beneath and will surface at some point. We believe we have reached this point and it is time to start building a new, general purpose serverless model — compatible with FaaS but designed from scratch. And yes, you may say we’re reaching for the moon, a small European start-up, we should know our place and so on…
But at Scramjet we weren’t listening. Sometimes all it takes is not to know what you are trying to achieve is impossible. And now we seem to have cracked it. And now we seem to have people strongly believing in our solution.
Serverless is a visible part of your cloud expenses and your teams spent weeks on looking for solutions to seemingly simple problems? Become an early adopter of Scramjet Cloud Platform a totally new serverless model that allows efficient processing of unlimited size of data in long lived processes and with microsecond-grade latencies
Feel free to reach us.