Run Live Energy Trading, Risk, and Analytics in Databricks

See how to combine live trading operations, risk monitoring, machine learning, and historical analytics into a single Data + AI Platform. Energy markets are becoming more volatile, while shorter settlement intervals increase the volume of records, orders, and risk recalculations. At the same time, trading desks operate against millisecond SLAs. When live trading operations and historical analysis run on separate systems, for example, an operational database for the trading desk and a separate platform for risk backtesting, machine learning, and regulatory reporting, teams face stale risk data, mismatched versions of the truth, fragmented governance, and costly synchronization pipelines. In this demo, presented by Victoria Sicking, Solutions Architect at Databricks, you’ll see a Databricks App presenting a live trading desk with prices, profit and loss (P&L), positions, risk, alerts, and price forecasts updating as the market moves. You’ll also see how market feeds flow through Spark Structured Streaming Real-Time Mode into Lakebase, a fully managed Postgres database native to the Databricks Data + AI platform. Lakebase serves as the OLTP system for millisecond reads, transactions, and upserts, with data governed by Unity Catalog. The demo also shows how machine learning models support trade signals and how analysts use dashboards and Genie to query historical trading data. What you’ll learn: • How to stream market data with Spark Structured Streaming Real-Time Mode into Lakebase for millisecond end to end latency from source to app and trade-signal. • How to monitor live prices, profit and loss (P&L), positions, and risk against desk limits – and trigger risk alerts as limits are approached. • How to train and deploy machine learning models on the same data to produce real-time price forecasts and trade signals. • How to explore historical trading data with embedded dashboards and ask Genie natural language questions about counterparties, trades, and buy-versus-sell volume. • How one data copy can support the app, analytics, and machine learning while Unity Catalog maintains governance, lineage, permissions, and audit end to end. Looking to get started? Getting Started with Apache Spark Structured Streaming Real-Time Mode - https://www.databricks.com/resources/demos/videos/apache-spark-rtm Connect a custom Databricks app to Lakebase -https://docs.databricks.com/aws/en/oltp/projects/tutorial-databricks-apps-autoscaling