Harnessing real-time data and machine learning, this startup wants to change how business intelligence works for you
Israeli incident detection and analytics startup Anodot announced on Thursday the close of their Series B funding round, pulling in $8 million in new capital.
Leading the charge was Aleph Venture Capital, with previous investor Disruptive Technologies L.P. participating as well. Combined with the money from their previous rounds, this new cash brings them to a total of $12.5 million raised overall.
Co-founded in June 2014 by CEO David Drai, Chief Data Scientist Ira Cohen, and VP R&D Shay Lang, Anodot works with clients to help get the most out of their data in making decisions that affect their business.
Offering a SaaS product, the company uses machine learning to understand a variety of data dimensions, building an understanding of what is supposed to be “normal” behavior, whether it be site traffic, development team operations, or perhaps most importantly to companies, changes in sales.
If, for example, a company sees a sudden spike or drop in their sales activity, it can signal that there may be a technical issue affecting their sales flow, preventing customers from making purchases. If this anomaly goes unnoticed for a week, it could severely impact a company’s revenues.
In the event that abnormality is discovered, it sends an alert to the team for closer inspection. Their reports to users examine this data, and put it into the context of what they call a bigger story, correlating it with other data to provide truly valuable insights.
Speaking with Drai, he tells Geektime that in the current state of the industry that does not work with real-time data, BI professionals are dependent on older data that is collected over time and only analyzed later when doing reviews.
Anodot believes that beyond their ability to provide up-to-date actionable insights to their customers, they stand out in their capacity to look at all the different elements of a client’s data, and provide precise insights relevant to specific teams.
“If you do an analogy to the BI world, they provide a platform, asking a question about a specific business,” Drai explains. “In our approach, we are asking the question for you with the machine learning that we have developed, and you only have to select which answers are interesting for you. Everyone can look through the answers that they provide, whether it is dealing with revenues, latency, or the machines. You select the topic that you are interested in, and we provide a subset of answers to meet your needs.”
“It is very important that they are agnostic to the dimensions” Drai tells Geektime, referring to the range of data sets being studied by their product. “Different companies have different needs from the data. We have customers from the IoT world, whose data can be from the machines building the products or from the users. Everything is different in terms of context for the client. We consider our technology as agnostic so that we can serve a wider customer base.”
Having written a fair amount on behavioral analysis in the cyber security space, this concept of looking for anomalies sounded familiar. Drai made a point of drawing out the differences from his product and that of many security companies, citing that a cyber researcher is looking for malicious individuals performing actions that break protocol, going where they are not meant to be. With Anodot, he says that they are looking for outliers in a system wide collection of aggregated data, looking for trends to report.
Conceptually, the two types of uses of behavioral analysis are similar enough for a basic understanding of how their autonomous BI tool works, even if their results end up in different places. Thinking about the problems with behavioral analysis, the most common complaint is the number of false positives that it can pump out, becoming a distraction for the professional using the product.
Acknowledging the concern, Drai says that they have algorithms that help to take care of the false positives. “First of all we have unique scoring on the behavior,” he explains about the settings for the alerts. The client gets to choose the sensitivity of the changes, receiving insights. To try and get this right, they perform simulations to fine tune what kind of insights the user wants to receive, allowing the same user to select various sensitivities for alerts of different KPIs.
Headquartered in Ra’anana, Israel, with a sales office in Santa Clara, California, Anodot has been selling their product since January, having pushed their services out to market in under two years. They have already managed to rack up an impressive 30 or so clients, with big names like Microsoft, Rubicon, Credit Karma, and Wix using them to make smarter business decisions.
Since receiving funding, they have nearly doubled their staff, growing to a team of nearly 25. Moving forward, they intend to continue hiring for their R&D and marketing teams, helping them reach their goal of working with more large enterprises.