February 2024 Summaries

Engineering teams must consider two key factors when deciding how to report usage records: volume of usage data and use cases that require aggregated usage. Usage-based pricing models require accurate recording of billable events, which can vary depending on the business, customers, and pricing model. Reporting directly to a database or using a dedicated metering system like OpenMeter depends on the scale of product consumption. Aggregated usage data is consumed in various functions across the company, including product, sales, and finance departments. Usage data use cases have different volume, latency, and integration requirements. Common alternatives for collecting and consuming usage data without a dedicated metering system include online databases, billing providers, OLAP systems, data warehouses, and streaming aggregation platforms. OpenMeter provides a robust solution for handling load, usage spikes, and duplicates required to implement accurate usage-based billing while seamlessly integrating with various organizational departments.

The OpenMeter team has introduced new features including subject metadata and display names, Postgres and ClickHouse collectors, webhook usage sync filters, and a LangChain example. Subject metadata and display names allow users to assign unique identifiers and descriptive names to subjects for easier identification in the UI and reports. The addition of Postgres and ClickHouse collectors enables users to import existing usage data from their databases into OpenMeter for further processing, reporting, and billing. Webhook usage sync filters now support filtering, allowing users to trigger notifications for specific usage events. Finally, an example demonstrates how to integrate OpenMeter with LangChain, a popular open-source framework for building language model applications.

The text discusses the importance of a tight feedback loop between customers and their product usage, emphasizing the need for real-time tracking to prevent unexpected costs and churn. OpenMeter's Consumer Portal and Web SDK enable businesses to offer real-time usage dashboards for their customers. By using React in three simple steps, companies can integrate OpenMeter's features into their products, allowing users to monitor consumption in real-time. The guide and example provided aim to assist developers in getting started with the Consumer Portal.

Companies must attribute shared costs to customers and internal teams for effective management of cost leadership. Finance and engineering should communicate using unit cost and internal rate cards, enabling efficient budgeting and forecasting. OpenMeter's FinOps unit economics solution helps companies attribute shared costs and predict vendor expenses like OpenAI token usage. By combining cost and meters into resources, businesses can break down resource costs by customers or groupings. Unit economics establishes a common language between teams and fosters better decision-making in product development and pricing strategies, ultimately leading to increased profitability.

The OpenMeter Kubernetes collector is a tool designed for accurately measuring pod execution time in Kubernetes environments. It simplifies per-second billing and metering by providing an out-of-box solution that can be deployed and configured using Helm charts. The collector uses Benthos, a robust stream processing tool, to collect and report information about running pods in the cluster. By default, it watches the default namespace and reports each pod's execution time every 15 seconds. However, users can fine-tune reporting intervals and namespaces according to their needs by referring to the documentation. The collector also comes with a few default configurations called presets that can be further customized using environment variables or directly editing the configuration file.

OpenMeter's native Stripe integration allows businesses to monetize various consumption data such as API requests, LLMs token usage, GPU and CPU time, and more. With this integration, users can quickly implement usage-based pricing for AI, API, and infrastructure consumptions like logs, Kubernetes, S3, and databases. The process involves three simple steps: starting metering with OpenMeter Cloud, connecting Stripe, mapping Stripe Customer IDs, and enjoying robust synchronization. This integration ensures precise billing by automating data transfer and synchronization processes, enabling businesses to focus on their core offerings while relying on OpenMeter and Stripe for seamless, hassle-free billing solutions.