UKOUG 2025: Just Streams: Real-Time Data Pipelines in OCI (with a Live Demo Twist) - Powered by AIDP
This year, I had the pleasure of joining the UKOUG Discover 2025 conference — a fantastic gathering of Oracle professionals, engineers, and data enthusiasts. Even more exciting, I had the opportunity to co-present with my colleague Sandi Holub, delivering what was (probably) one of the most hands-on sessions of the entire event.
Our session, titled Just Streams: Real-Time Data Pipelines in OCI (with a Live Demo Twist), focused on exactly that: no slides, no theoretical diagrams — just real data in motion.
A Session Built on Real-Time Everything
We set out to demonstrate what becomes possible when you combine streaming data, cloud-native services, and a bit of creativity. The entire session was built around real, public data from Transport for London (TfL) and showcased two live demonstrations:
- Real-Time Bus Tracking and Arrival Analysis
We ingested live bus arrival data into OCI Streaming, processed it with OCI Data Flow (Apache Spark), and pushed curated results into Oracle Analytics Cloud. The outcome: analysts could run SQL-based queries and visualize insights within seconds of events happening on the streets of London.
- Real-Time Bus Movement Visualization
The second demo focused on live GPS telemetry from TfL buses. Using GoldenGate Stream Analytics, we enriched streaming location data with route context and displayed it on a real-time dashboard — every location update brought the visualization to life.
Although our examples came from public transport, the same architecture applies to real-world industry scenarios — from manufacturing plants and steel mills to logistics, IoT networks, and sensor-driven operations. We even prepared a small ML/AI “cookie” — showing how anomaly detection or predictive insights could be layered on top of streaming pipelines — though we didn’t include it in the final demo.
A Last-Minute Twist: Rebuilding Everything on Oracle’s AI Data Platform
One month before Discover 2025, during AI World in Las Vegas, Oracle announced the Oracle AI Data Platform (AIDP) — a unified, AI-native architecture designed to bring together streaming, data engineering, machine learning, and analytics into a single, coherent ecosystem.
During a quick Teams discussion, Sandi and I looked at each other and said: “We’re doing this on AIDP.”
And so, at the very last minute, we rebuilt our entire demo environment on top of Oracle’s brand-new platform.
The Result?
A fully working real-time streaming pipeline running natively on AIDP — feeding directly into Oracle Analytics for live reporting.
By the time we arrived at UKOUG, our pipelines weren’t just running — they were performing better, cleaner, and more efficiently thanks to AIDP’s integrated, AI-native design.
A Few Words on the Oracle AI Data Platform (AIDP)
AIDP represents a major shift in how data and AI workloads are built on Oracle Cloud. It brings together:
- High-throughput streaming and ingestion
- Unified lakehouse storage
- Distributed compute for Spark, Python, and AI workloads
- Built-in vector search and model serving
- Seamless integration with Oracle Analytics, GoldenGate Stream Analytics, and the broader OCI ecosystem
For teams building real-time, AI-driven data pipelines, AIDP eliminates fragmentation. Everything — from ingestion to enrichment to machine learning to analytics — connects through a single, AI-native fabric.
Our experience at UKOUG showcased just how powerful that can be.
A Conference to Remember
Presenting at UKOUG Discover 2025 was a fantastic experience, and doing it alongside Sandi made it even better. The energy in the room, the number of questions, and the excitement around real-time architectures and the new AI Data Platform were truly motivating.
With Oracle doubling down on AI-native data architectures, I’m confident this is just the beginning. And yes — we’re already preparing the next demo.
Download slides here.