Kinetica is a real-time analytic database company ideal for sensor and machine data workloads. And Kinetica leverages GPUs and modern many-core CPUs to achieve robust performance and efficiencies. Pulse 2.0 interviewed Kinetica co-founder and CEO Nima Negahban to learn more.
Formation Of Kinetica
In 2009, Nima Negahban and Kinetica co-founder Amit Vij were contracted by the United States Army Intelligence and Security Command and the National Security Agency (NSA) to create a system to track and capture terrorists in real time.
“No database existed in the market to meet the NSA’s demands. The emerging big data technology back then was focused on batch analysis of web data. The mission called for real-time analysis of a more diverse set of big data sources,” said Negahban. “So we built a new class of enterprise-grade SQL database from the ground up that could provide instant results and the ability to visualize insights across massive, spatial and time-series streaming datasets. At the time, our database ingested data from more than 200 different streaming big data feeds. This included drones that tracked every asset that moved at 30 frames per second; mobile devices that emitted their metadata every few seconds; social media like Twitter and Facebook; and cyber security data. We were evaluating billions of signals to find that needle in the haystack. We received our first patent in 2011 and deployed at the US Army Intelligence and Security Command in 2012. Our first commercialized product launched at the United States Postal Service in 2014. Former President & COO of Oracle, Ray Lane contributed funding in 2016, and Kinetica opened its offices in San Francisco, and adoption started to grow the business.”
Favorite Memory
What has been Negahban’s favorite memory working for Kinetica so far? “One of my favorite memories was watching the first queries run against Kinetica on a small 1-u single GPU-based machine far outpacing the huge clusters our research program used to perform the same queries. That was the point I think we realized our idea of creating a fully optimized database to leverage many-core compute devices like the GPU was going to have a serious future,” Negahban shared.
Challenges Faced
Has the current macroeconomic climate affected the company? “Prior to the start of Fed interest rate hikes, Kinetica was focused on getting lean and achieving financially responsible growth. At that time, I’d hear from venture capitalists who would say they prefer we were focused on growth exclusively at all cost,” Negahban acknowledged. “Once the reality set in with start-ups and VCs that funding was scarce, we were ahead of the game. In fact, I recently met with a bank who said they hadn’t seen a combination of growth and expense management in a startup like ours in a long time.”
Core Products
What are Kinetica’s core products and features? “Kinetica’s design center is fusing massive geospatial and time-series data sets together and processing complex spatiotemporal analytics in real-time. Kinetica uses native vectorization to outperform other cloud analytic databases significantly,” Negahban explained. “In a vectorized query engine, data is stored in fixed-size blocks called vectors, and query operations are performed on these vectors in parallel rather than on individual data elements. This allows the query engine to process multiple data elements simultaneously, resulting in faster query execution and improved performance, particularly those that require fusion of temporal and spatial data in real-time.”
Evolution Of Kinetica’s Technology
How has Kinetica’s technology evolved since launching? Negahban noted:
Kinetica’s technology has evolved with the proliferation of connected devices and sensors that have advanced from taking readings over time to taking readings over space and time. Understanding this trend and the resulting impacts is essential for innovators seeking to create value in the next wave of IoT products and services. Prime examples are data streams from mobile devices, static or moving sensors, satellites, and video feeds from drones and closed-circuit TVs. Conventional analytic databases were designed to analyze transactions and first-generation Big Data like weblogs. But, getting value from sensor data characterized by time-stamps and geo-encoding requires new capabilities that aren’t satisfactorily addressed by prior-generation databases, even those with special object-relational extensions for spatiotemporal data.
Kinetica is focused on fusing massive geospatial and time-series data sets together and processing complex spatiotemporal analytics in real-time. Many customers refer to their Kinetica implementation as the ‘speed layer.’ It fills a critical gap between traditional data warehouses and data lakes that are batch-oriented and optimized for transactions and log data and streaming tools like Kafka, Confluent, and Kinesis that are real-time but unable to run advanced analytics. Kinetica is available as-a-Service, and everything from connecting to Kafka queues to invoking powerful spatial and time series functions is done through simple SQL, which translates into incredible time to-value. Now that real-time spatial data is proliferating across so many industry sectors, Kinetica is a proven and hardened database that is accelerating the adoption of real-time location intelligence.
Significant Milestones
What have been some of Kinetica’s most significant milestones? “Kinetica was first deployed at the US Army Intelligence and Security Command in 2014 and commercialized and launched at the United States Postal Service in 2014. Kinetica’s deployments have been recognized with two prestigious IDC awards for ‘High-Performance Computing Innovation Excellence’ in 2014 and 2016,” Negahban highlighted. “In 2019, Kinetica introduced the new Active Analytics platform, with improved AI and ML integration, a graph server for routing, automated installation, and configuration in a high-performance, cloud-ready package. Telecoms, security, retail, transportation companies, and more are now using Kinetica. In 2021 Kinetica became available as a fully managed service in the Cloud. Continued focus on core features and an improved Workbench interface drives growth among enterprises, new tech, and government agencies, including some of the world’s largest IoT data analytics implementations. In 2023 Kinetica announced Conversational Query and became the first OpenAI ChatGPT integration with an analytic database.”
Customer Success Story
When I asked Negahban about a customer success story, he cited the FAA, Citi, Liberty Mutual, and T-Mobile. “The FAA uses Kinetica for Air Domain Analytics by fusing hundreds of feeds in real-time for more accurate and faster monitoring of our air space. Citi uses Kinetica for Real-time Trading by constantly recalculating complex financial metrics in near real-time. T-Mobile uses Kinetica for Network Optimization that improves 5G coverage resulting in better cell service for customers. Liberty Mutual uses Kinetica to modernize Claims Management by fusing real-time weather events with insured building footprints resulting in more accurate and timely responses to catastrophic weather events. Real-time location data exists across multiple industries and is enabling all sorts of new high-impact outcomes,” Negahban commented.
Funding/Revenue
In February 2023, Kinetica announced record business momentum over the past year of 90% Annual Recurring Revenue (ARR) growth, Net Dollar Retention Rate (NDRR) of 163%, and a doubling of its customer base. Leading organizations creating the next generation of data-driven applications based on sensor and machine data continue to choose Kinetica for its unrivaled real-time analytical and processing power of time-series and geospatial data. To date, Kinetica has received $77 million in venture capital funding,” Negahban revealed.
Total Addressable Market
What total addressable market (TAM) size is Kinetica pursuing? “Deloitte estimates that devices capable of sharing their location will represent 40% of all data by 2025, making spatiotemporal data – where objects are and where they are moving – the fastest growing segment of big data. Prime examples are data streams from mobile devices, static or moving sensors, satellites, and video feeds from drones and closed-circuit TVs,” Negahban assessed. “Applications based on real-time spatial and time-series data are driving digital transformation across industries, including fleet management, supply chain transparency, connected car, precision agriculture, smart energy management, retail proximity marketing, and many others.”
Differentiation From The Competition
What differentiates Kinetica from its competition? “As stated earlier, Kinetica uses native vectorization to outperform other cloud analytic databases significantly. In a vectorized query engine, data is stored in fixed-size blocks called vectors, and query operations are performed on these vectors in parallel rather than on individual data elements. This allows the query engine to process multiple data elements simultaneously, resulting in faster query execution and improved performance,” Negahban affirmed. “Recent independently verified industry-standard benchmarks show Kinetica is 5X faster than Google BigQuery, and 12X faster than ClickHouse. This leap in performance allows Kinetica to address previously intractable workloads, particularly those that require the fusion of temporal and spatial data in real-time. Kinetica supports dozens of spatial and temporal join types, hundreds of in-database analytic SQL functions, and enables visualization of billions of data points on a map.”
Future Company Goals
What are some of Kinetica’s future company goals? “The emergence of generative AI is very interesting to us and our clients. Kinetica was the first analytic database to launch a ChatGPT interface to turn language into SQL. This has been extremely well received, particularly since Kinetica’s architecture is ideally suited for rapid answers to ad-hoc questions, which make querying more conversational,” Negahban concluded. “But that’s just step one in how generative AI will influence our roadmap and capabilities. Ultimately, our goal is to be the de-facto platform for enterprise data analytics based on generative AI.”
