EDA

Enterprise Data Analytics (EDA) is a full‑stack, self‑hosted analytics platform designed to bridge the gap between complex data warehouses and end‑user dashboards. At its core, EDA exposes a web application that lets users craft SQL or “tree‑mode” queries against a MongoDB data store, then visualise the results in drag‑and‑drop dashboards. For developers, this translates into a RESTful API layer written in Node.js, a front‑end built with Angular (or React, depending on the fork), and a persistent layer that leverages MongoDB’s flexible schema to store both metadata (data models, dashboards, user roles) and runtime artefacts (query results, KPI thresholds).

• Backend – Node.js (v18+), Express‑style routing, TypeScript for type safety. The API layer exposes endpoints for CRUD operations on data models, dashboards, KPIs, and user permissions. It also hosts a WebSocket hub for real‑time KPI alerts.
• Frontend – Angular (or an equivalent SPA framework). The UI consumes the API, renders dashboards with a lightweight charting library (e.g., Chart.js or Highcharts), and offers an interactive query builder that can toggle between raw SQL and a visual “tree mode”.
• Database – MongoDB (local or Atlas). All configuration is stored in database.config.js, and the database contains collections for users, roles, dashboards, queries, and audit logs. Row‑level security is implemented via MongoDB’s document filtering.
• Containerization – A single Docker image (jortilles/eda:latest) bundles the API and front‑end, exposing port 80. Kubernetes manifests or Helm charts are available for orchestrated deployments.

• SQL & Tree‑Mode Querying – Developers can embed raw SQL or use the visual builder; both generate identical MongoDB queries under the hood.
• KPI Definition & Alerts – APIs to create thresholds and trigger email or webhook notifications when metrics cross bounds.
• Public Dashboards & Embedding – Each dashboard gets a shareable URL; embedding is achieved via an iframe with optional authentication tokens.
• Parent‑Child Reports – Hierarchical report relationships are persisted in MongoDB, enabling drill‑down navigation.
• Row‑Level Security – Role‑based filters are applied at query time, ensuring users only see data they’re permitted to view.

EDA is intentionally lightweight: a single Docker container runs on any Linux host with Docker, or can be deployed to cloud providers via one‑click buttons (AWS CloudFormation, DigitalOcean App Platform, Render, or Helm). The only external dependency is a MongoDB instance; developers can run it locally with Docker Compose or connect to a managed cluster. Horizontal scaling is achieved by running multiple API instances behind a load balancer, with dashboards served statically from the same container or via an external CDN.

• Plugin Hooks – The API exposes lifecycle hooks that allow custom modules to register new chart types or data connectors.
• Webhooks – KPI alerts can be routed to Slack, Teams, or custom endpoints.
• SDKs & Client Libraries – While not bundled, the REST API is fully documented (OpenAPI spec) and can be consumed by any language with HTTP support.
• Custom SQL Dialects – Developers can extend the query parser to support additional database backends beyond MongoDB.

Configuration is straightforward: edit database.config.js for the DB URL and config.ts for API endpoints. The codebase follows TypeScript conventions, making IDE auto‑completion a breeze. Documentation is concise but covers all public endpoints and data model schemas. Community support is modest; the GitHub repo hosts issue trackers, but there are no dedicated forums or Slack channels. Nonetheless, the open‑source nature allows developers to fork and modify without licensing constraints.

• Internal BI Tool – Companies that need a quick, self‑hosted analytics layer without the cost of SaaS BI platforms.
• Data‑Driven Applications – Embed dashboards directly into existing web apps to give users real‑time insights.
• Compliance & Auditing – Row‑level security and audit logs make it suitable for regulated industries that require strict data access controls.
• Rapid Prototyping – Teams can spin up a Docker container in minutes, define data models via SQL, and iterate on visualizations without deploying a full‑blown ETL pipeline.

• Performance – MongoDB’s indexing and Node.js event loop provide low‑latency query execution for typical BI workloads.
• Flexibility – Developers can tweak the source code, add new chart types, or replace MongoDB with another NoSQL store if needed.
• Licensing – Completely open source under an MIT‑style license; no subscription fees for self‑hosted deployments.
• Simplicity – The UI abstracts away complex BI concepts, letting developers focus on data modeling rather than UI scaffolding.

In sum, EDA offers a developer‑friendly, self‑hosted analytics stack that balances ease of use with the ability to extend and integrate into larger ecosystems.