Data extraction co Rookout raises $8m

Or Weis / Photo: Rookout

The Israeli company's product makes it possible to get information from live code while it is still running.

Israeli startup Rookout , which is developing software for extracting data from live code, today announced an $8 million financing round. Cisco Investments, the investment arm of Cisco Systems, led the round, with participation from Rookout's previous investors: TLV Partners, the Emerge fund, and GitHub CEO Nat Friedman. The new financing round brings the total raised by the company to over $12 million.

Rookout's founders, CEO Or Weis and CTO Liran Haimovitch, served together in the IDF Intelligence Corps. They founded the company in 2017 in order to solve a problem that troubled them as developers in the army and in the civilian market. Rookout has 20 employees, including 18 in its offices in Tel Aviv and San Francisco. The company plans to double its staff in the coming year, mostly by hiring sales and marketing staff in the US.

Processes of finding information in code in order to detect and correct errors, for example, usually begin at the code-writing stage. The programmers write stopping points into the code - conditions that determine when the code will stop running, so that the programmer can analyze the program and make necessary changes - but the applications is suspended while this is taking place. Rookout has developed software designed to enable engineers to search for and find information in live code without halting its running, and without writing logs. A log is code written separately for the rows of code in the application in order to retrieve information from each row for which it was written, while the application is running.

"The unique aspect of our technology is in the decoupling of data from code, which gives access to the app's data layer without affecting the code layer," Weis explains. Rookout's product operates as a separate directory that is capable of reading the information of the live application while it is running on the server's memory. When you write a command in this directory, it is capable of retrieving information from the live application running on the server's memory by scanning the application and automatically creating "stopping points that do not stop" according to the command inserted by the programmer. At these points, the system gathers information about the reason for the halt and how the code is running, and sends it to the programmers without stopping the live code. In contrast to the stopping points that have to be inserted in advance, Rookout's stopping points operate in real time and on demand.

"The software sector is also exploding in both the rate at which we are creating software and its volume. They're developing huge projects today composed of tens of thousands, and sometimes millions, of parts that are constantly and dynamically changing," Weis told "Globes." "Our solution really changes the way people look at software. People today are trying to log in to everything all of the time, because they are afraid that they won't manage to write a log when it is needed. It requires a lot of resources and generates a lot of garbage. Using our product enables programmers to change what they're doing at any time, and to focus more on developing the product and less on what they need to keep track of."

Weis adds that although the company began by developing a solution for the basic use of detecting errors in the code, the product has now been expanded to additional uses at the request of the customers. Today, the product supports various cloud environments, including serverless applications in the Python, NodeJS, and JVM languages. Rookout plans to use the money it raised to expand support of its product and to additional programming languages and code environments.

"We discovered that data analysts are in a worse position than the programmers. In order to gather data from machine learning algorithms, they have to ask the developer for the logs, so their work circle is even longer and more complicated than that of the programmer," Weis says. "Our product enables the data scientist to look into the algorithm and see how it develops and learns. It's like a tree that's growing. If you look at it only after it grew, it will be very difficult to understand how it got to where it is. Looking at it during the process enables a person to look, see, and analyze while the process is taking place.

Published by Globes, Israel business news - en.globes.co.il - on August 7, 2019

© Copyright of Globes Publisher Itonut (1983) Ltd. 2019

Or Weis / Photo: Rookout
Or Weis / Photo: Rookout
Twitter Facebook Linkedin RSS Newsletters גלובס Israel Business Conference 2018