Under the Hood of OpenCourtData.uk: Our Serverless, Agent-Driven Tech Stack

When building a civic technology platform dedicated to the programmatic aggregation of UK public records, speed, scale, and reliability are paramount. Extracting, processing, and serving vast amounts of statutory transparency data and court judgments requires an architecture that won't buckle under pressure.

Here is a deep dive into the serverless and AI-driven tech stack that powers OpenCourtData.uk, enabling rapid development and edge-optimized data delivery.

1. The Cloudflare Ecosystem: Compute and Data at the Edge

The foundation of the platform relies entirely on Cloudflare's serverless ecosystem, allowing us to keep latencies low and deployments highly automated without managing traditional infrastructure.

Cloudflare Pages

We host the frontend on Pages, ensuring the platform is globally distributed and lightning-fast for users searching through council registers or judgments.

Cloudflare D1

For our database layer, D1 provides a native, serverless SQLite experience built right into the edge. It seamlessly handles the relational data structures required for organizing complex public records.

Cloudflare Workers

The heavy lifting of our data extraction pipelines happens here. Using Workers, we deploy highly analytical, automated crawlers that gather and aggregate data. Developing these edge-based agents often involves Python-in-Workers, managing dependencies with tools like uv, and leveraging Pyodide to run smoothly in a WebAssembly environment.

2. The Agentic Development Workflow

Writing and maintaining the codebase for a data-intensive platform has fundamentally shifted with the introduction of autonomous AI coding agents.

Google Antigravity

This multi-agent IDE serves as our mission control. By utilizing Antigravity's agent-first architecture, we can orchestrate complex full-stack scaffolds and rapid UI iterations. Its specialized file system and browser agents help verify code behavior on the fly, allowing us to focus on the overarching architecture rather than writing boilerplate.

Google Jules & GitHub

Jules acts as an asynchronous developer operating directly within our GitHub repository. When an issue is tracked and scoped out—often moving straight from our Linear issue tracker to execution—Jules spins up a secure cloud VM, writes the necessary tests and features, and opens a pull request. It is the perfect loop for verifying high-severity security patches or bumping dependencies without breaking flow state.

3. Analytics and Monitoring

Understanding how users interact with civic data is crucial for continuous improvement and maintaining a strong platform strategy.

PostHog

Gives us granular, event-driven product analytics to see exactly how search features and data filters are being utilized.

Google Analytics & Search Console

These form the backbone of our SEO optimization, tracking acquisition channels and ensuring that public records index properly so they can be easily discovered by the public.

Infrastructure Telemetry

In the background, we pair this product data with deep infrastructure monitoring—running custom NRQL queries via New Relic agents to track CPU and memory usage across our fleet of crawler VMs.

4. User Engagement and Authentication

Accessibility extends beyond fast queries; it's also about a frictionless and secure user experience.

OAuth Authentication

We implemented Google and Microsoft OAuth, allowing legal professionals, researchers, and citizens to log in securely with single sign-on (SSO). This completely bypasses the friction and security risks of standard password creation.

Brevo

To keep our community engaged, Brevo handles our automated email marketing campaigns. This handles our multi-stage welcome sequences that onboard new users, guiding them through the platform’s advanced search capabilities and data export tools.

5. Lean, Scalable Civic Technology

By combining the raw edge-compute power of Cloudflare with the massive velocity gains of Google's agentic development tools, OpenCourtData.uk remains lean, scalable, and entirely focused on its mission: making public data truly accessible.