Why we invested in synnada.ai
A fundamental trend of the past decade is that every company is becoming a software company and a data company. This shift resulted in orders of magnitude more data in a way less structured form and significantly more urgency to process and consume them as a stream (opposed to batches, the previously dominant paradigm in data infrastructures). The world moved away from quasi-static big data to dynamic, behavioral stream data. Which, besides analytics, has been generating value for operational purposes and has become mission-critical, deeply embedded in almost all processes and applications of modern enterprises.
This continuous data flow is the backbone of many complex automated systems, from social networks to production lines. However, it doesn’t allow analysts to sit back and comfortably dive clean and analyse each data set, as an abundance of information is generated every millisecond. Instead, the goal is to observe a system, detect certain behavior and take necessary and relevant actions.
Data observability is all about eliminating downtime, or in other words, enabling the identification, troubleshooting and resolving of issues in near real-time.
Detecting behavioral issues is typically done against a declarative set of rules in the form of alerts. The problem is that oftentimes these rules are static. As time goes by and the datasets’ environment changes, the rules become obsolete and inaccurate without manually recalibrating them. In the end, engineers end up with far more data points flagged outside the threshold than actually problematic ones (false positives). Moreover, the manual investigation of these keeps whole engineering teams busy every day.
What if we could have better models? Maybe even dynamic ones?
Synnada is building a self-contained machine learning model environment that, besides discovering alerts and monitoring them, but also learns from the outcome of these alerts to improve its own detection model. With continuous accurate operationalization, the product offers a cost-effective way for companies of all sizes to have more precise data insights and focus on what matters.
Ozan is an AI scientist with solid academic (ODTU, Stanford) and professional (Optumsoft, Striim, and Facebook) backgrounds and built event detection systems in media, finance and industrial verticals.
Sam ran operations and grew teams from 20 to 100 people & helped raise $35M+ in funding for Apsiyon & Picus. He also co-hosts the leading local podcast on entrepreneurship in Turkey.
We are thrilled to have joined the company’s 2.6m USD pre-seed round with 500k USD to support Ozan and Sam in making data tools as sopisticated as software engineering tools.
