Sapient Industries logoCase study

Sapient uses QuestDB for real-time analytics in smart infrastructure

QuestDB is used by Sapient Industries in their proprietary analytics platform, which modernizes building infrastructure by monitoring energy data from electrical equipment for usage insights that enable anomaly detection, energy savings, asset management, and facility optimization.

An illustration showing some of the capabilities of Sapient Industries' analytics platform

Dollar iconControl costs by using fewer cloud resources to reach ingestion speeds

Workflow iconCompatible with RESTful components, PostgreSQL tooling

Leaf iconSQL queries are simple to read and intuitive to compose

Gauge iconAnalytics platform for efficient ingestion of hundreds of GB/month

Voice iconSimple setup with minimal maintenance

Time iconHigh-performance solution for real-time analytics

Sapient empowers customers to unlock and leverage real-time electrical data to solve the increasing challenge of efficient facility operations. The insights derived from these analytics help customers build more sustainable and efficient operations to manage infrastructure, predict maintenance cycles, and reduce energy costs.

In this case study, Dan Worth, Principal Software Engineer at Sapient Industries describes how and why QuestDB is relied upon within high-performance clusters for real-time analytics.

Sapient Industries provide analytics for smart buildings

Our integrated platform applies machine learning and analytics to facility electrical data, delivering deep insights that drive cost, carbon, and operational efficiencies. Our customers capitalize on next-level building intelligence to harmonize and execute energy, facilities, sustainability, and real estate initiatives.

Our solution supports stakeholders across facilities management, energy, real estate, sustainability, asset management, IT, and finance. Our platform’s analytics drive immediate impact and intuitively enable data-driven decision-making by providing users with actionable insights at every node of a building’s energy consumption.

A diagram showing how Sapient Industries collects and consolidates input to contextualize industrial sensor data

Why we use QuestDB for real-time analytics

QuestDB facilitates actionable performance because of the speed for complex queries. Data can be pulled in real-time without having to store it. One major draw for us was that QuestDB uses SQL; this eliminated the learning curve for our team. We don’t have to learn a new custom query language and we can easily compose common queries as we go with our existing familiarity with SQL.

We’re capturing minute-level energy usage data from electrical equipment throughout facilities. These devices and pieces of equipment range from IT gear like laptops and workstations, to industrial machinery, lighting, facility equipment, refrigerators, and all other energy-consuming equipment.

Upon collecting electrical consumption data from the thousands of pieces of equipment in a building, we use machine learning algorithms to automatically label each device throughout the facility. We then employ a collection of ‘device-utilization’ queries to check for abnormalities such as sudden amperage increases, deliver energy usage comparisons over time and geography, provide emissions calculations, among countless other metrics.

When customers have a detailed breakdown of the equipment-specific energy consumption in their facilities, they can make data-driven decisions that have broad-reaching benefits across the organization. Insights driven by the detection of energy usage spikes, abnormally high energy density, or irregularities in current draw drive efficiencies across maintenance, space reconfiguration, procurement decisions, sustainability investments and more.

Our analytics platform is cloud-native and highly available

For our deployment, QuestDB is configured as a StatefulSet in a Kubernetes cluster on Google Cloud Platform. We deploy two instances on different node-pools and route traffic to both for High Availability. We back up partitions hourly to Google Cloud Storage using custom Go programs that we developed.

Our deployment’s API queries QuestDB over REST API, and we’re ingesting data using InfluxDB Line Protocol for high throughput. For the data requirements, we’re typically ingesting 250GB worth of sensor data into QuestDB per month, and we currently have about 1.2TB in an instance.

Sapient enables building intelligence by providing the data needed to make smart and proactive operational decisions, which yield multiple larger benefits around facility planning, asset investments, and sustainability goals. Our proprietary software generates these data insights, and we chose QuestDB for the simplicity of SQL, high-throughput ingestion, and the compatibility of our tooling with PostgreSQL wire.

Dan Worth, Principal Software Engineer at Sapient Industries