The 10x Difference

Posted on: . by Pradeep Madhavarapu

Every once in a while, new technology at one layer enables a whole host of products at another layer to be reinvented, providing new features or solving new use cases that could not be solved earlier. I was fortunate enough to build a breakthrough product, Amazon Aurora, during one such episode of reinventions – the elasticity and multi-tenancy of cloud computing plus the availability of SSDs allowed us to build a practical distributed relational database that was 10x better than the status quo. When Darshan and Monish talked to me about Isima, I was pleasantly surprised to see a similar opportunity for reinvention in the big-data and analytics space.

It Doesn’t Make Any Noise!

In the ’80s, the first generation of refrigerators had been adopted by the Indian middle class, and the second generation was just coming out. One of the players had a TV ad that stayed with me1. In the ad, a new owner of the refrigerator calls a technician to repair it. The technician checks it out, doesn’t find anything wrong, and asks whether it keeps things cool, makes ice, etc. The owner says yes, it keeps things cool the way they expect it to. So the technician asks why they called him, and the owner responds, “It doesn’t make any noise!”2 Refrigerators were expected to make a lot of noise, and when this one did not3, the owner thought it must be malfunctioning. This is what people experience when a 10x improvement comes along once in a while.

Utility for a Few vs. Usability for Everybody

Most of the big data platforms prevalent today are based on projects that were built for internal use by tech titans like Google, Facebook, Yahoo, and LinkedIn. These companies have a vast number of complex use-cases, the largest scale of data, and the best engineers in the world. They created these tools and optimized for utility for themselves rather than usability for a broad range of companies. Consequently, most companies outside this league of large tech companies find it difficult to use these tools and get bogged down by the complexity involved in building and maintaining large architectures with a medley of mismatched systems. Contrast that with the previous generation of data platforms like relational databases, file servers, and storage systems. They were designed to be sold to a wide variety of companies and were successfully used with much less frustration.

At Isima, we are building such a big data platform that is designed to be easy to use. It is not an incremental improvement to existing systems; rather, it is a re-imagination of what a big data platform should be like. It spans the whole lifecycle of data management, including ingestion, storage, processing, visualization, and utilization. This allows customers to build their entire analytics application on one system without having to cobble together many disparate systems. Our goal is to delight our customers with an experience that echoes the sentiment of “it doesn’t make any noise!”

If you want to build such a delightful product, and you are an expert in databases or distributed systems, I would love to talk to you (we are hiring)!

To hear perspectives from my fellow partners in crime, see here, here, and here.  The press release announcing our launch is here.

To get started on your journey, click here. And you can reach us @ or +1(833)-API-AIBI.

[1] I don’t remember the ad fully, and could not find a link to it; please let me know if you find it.
[2] In Hindi it was “awaaz hi nahin karta.”
[3] Compressor technology had advanced enough to allow refrigerators to be significantly quieter. One could say this was the same kind of advancement Tesla brought to cars.