The PingThings platform is a petabyte-scale, high performance platform for managing, analyzing, driving applications with and applying machine and deep learning to time series data to turn sensor data into an asset.

The company covers numerous verticals:

• INGEST - Supports a wide range of grid sensors including SCADA, DFR, AMI, and Smart Meter, synchrophasor, distribution synchrophasor, continuous point-on-wave, power quality, and more.
• STORE - Can persist, access, and query sensor data at up to 1 billion samples/second/stream. Also allows users to capture and access supporting data for grid analysis, topology and velocities of Time Series Data, Network Topology, and Sensor Metadata
• ANALYZE - Ad-Hoc Analysis (rapid prototyping via Jupyter Notebooks and a range of languages) and Large-Scale Data Processing (proprietary DISTILTM framework allows high speed time series analysis) and integration with open source frameworks such as Apache Spark, Ray, Google’s TensorFlow, etc.
• ACCESS - Both RESTful JSON access and high-performance binary APIs for numerous programming languages are available to complete with access control with on-prem or (preferred) cloud deployment (AWS/GovCloud or Azure).
• DRIVE APPLICATIONS - The platform can run and support dozens of simultaneous applications driven by sensor data both in real time or asynchronously.
• LEARN - train various ML and AI algorithms with time series data and then perform inferencing at scale.

SaaS business with cost based on compute and storage consumed.

• High Performance Time Series Database - Purpose-built time series database specifically for large numbers of sensor data streams and high frequency sensors. Can read and write tens of millions of points per second per node.
• Proprietary Data Structure - Invented a new data structure for machine-generated sensor data that supports instantaneous temporal aggregations, out-of-order data insertions, nanosecond time precision, dynamic sampling rates, and multi-resolution time series, and data versioning.
• Custom Analytics Framework - Built a horizontally scalable analytics framework optimized for computations and transformations applied to time series data at a scale designed to ease the complexities of big data.
• Resilient and Robust Infrastructure to Minimize Cost - Able to scale, compute, and storage layers to be an optimal performance to cost ratio and automatic data replication and self-healing for durability, high availability, and fault tolerance. The entire platform is containerized, and orchestrated using Kubernetes.
• Proprietary Data Compression - Created a new method to compress time series data to reduce storage costs, accelerate data access, and preserve all information captured by the sensors.

• Interactive Data Visualization - Human-scale visualization of arbitrary amounts of sensor data for real-time user data exploration
• Rapid Analytic Prototyping - Decrease cost of analytic experimentation and prototyping
• Real Time Event Detection - Detect anomalies and other events within time series data
• Additional use cases - cybersecurity, capital optimization, model verification and validation, predictive asset maintenance, process automation, rapid event analysis/reporting, and others