Designing Area Code. Fast Analytics Apps for modern devs.

Creating performant analytics apps today means stitching together dozens of complex systems — data ingestion, APIs, frontend frameworks, observability, orchestration. Until now.

Alex

4 min readAug 7, 2025

What is Area Code?

Data is hard. And often slow to work with. Area Code makes that easier. It’s a subtle nod to our name 514, which is the area code for Montreal. It’s our answer to this challenge: a monorepo starter designed for building modern, multi-modal applications, especially in the data and analytics space. It’s minimal, modular, and built with developer experience at the forefront. Devs often don’t know where to start with some of this stuff, so we decided to explore starter kits and templates.

No, it’s not a Ludacris song. Area Code is a Turborepo-based monorepo starter project that comes with batteries included. It’s a fully-integrated environment for developing production-level applications — particularly those centered around data.

Beautiful, Accessible, Fast Analytics Powered by the Best (including our own)

Out of the box, you get:

Pre-configured apps (frontend, backend, ingestion, analytics)
Integrated developer tooling with pnpm, Vite, and Turborepo
Modern architecture powered by TypeScript and React 19
Automated setup scripts and sample flows

Area Code is currently comprised of 2 flagship starter repos:

User Facing Analytics (UFA): A TypeScript-based, user-facing analytics stack that connects frontend, backend, and streaming pipelines using Moose Workflows.
Operational Data Warehouse (ODW): A Python-based ingestion and analysis stack, using Moose to pull from blob storage, logs, events, and more. The UI is built with Streamlit, tailored for internal data workflows.

Overview:

A real-time pipeline powered by Moose for ingestion and serving
A modern frontend stack: React 19, Tailwind, shadcn/ui, and TanStack (Router, Query, Form)
A responsive design system that’s both developer-friendly and enterprise-grade
AI Chat assistance (Vercel’s AI SDK)
High-performance data layers via ClickHouse and Elasticsearch
Robust customizable Foo/Bar Demo experience

UFA under the hood:

Transactional: PostgreSQL | Fastify | Drizzle ORM
Analytical: ClickHouse | Moose (API & Ingest)
Search: Elasticsearch

ODW under the hood:

Connectors, Frontend and Backend
ClickHouse (Analytical Database)
Redpanda (Message Queue)
Temporal (Workflow Engine)
Redis (Caching)

OpenSource, Agents, ShadCN Components, beyond

Real-world SaaS development brings a host of constraints, especially in enterprise contexts. Building Area Code meant solving for those from the start.

OpenSource: The monorepo structure works seamlessly with GitHub workflows, Cursor Rules, and Actions — great for teams operating in regulated or enterprise environments.
Agentic-friendly: Designed to auto-operate while integrating through shared contracts and workflow triggers. This enables cleaner boundaries between teams and services, and leverages an AI Chat SDK.
ShadCN Components: Using Vite, Shadcn other frontend frameworks, Area Code supports rapid iteration with accessibility and themeability baked in. Tailwind ensures consistent design without slowing you down. This makes Area Code a reliable foundation for real production teams.
Beyond. The idea was to validate and continue to build templates to better kick start common projects. This would allow faster entry and fewer steps to get started and offer a vercel-like cloud hosting service to users with little friction.