Uwazi Document Collections on AWS - Managed Support

This is a repackaged open source software product wherein additional charges apply for cloudimg support services.

## Build and Publish Structured Document Collections

Uwazi is an open source platform for organising, searching and publishing collections of documents and records. Human rights defenders, investigative journalists, legal teams and researchers use it to turn scattered evidence, case files and archives into a structured, fully searchable library. This image delivers Uwazi built and configured on a native Node.js application with dockerised MongoDB and Elasticsearch, so your team can define its first template and upload its first document within minutes of launch.

## Who This Is For

Human rights and legal organisations cataloguing case files, testimonies and evidence into structured records that relate to one another, with full text search across every uploaded PDF. Investigative and research teams building a private working archive or a curated public library from primary source documents. Librarians and archivists who need a self hosted collections platform inside their own VPC without spending ops cycles on installation, search configuration or credential rotation.

## What Makes Uwazi Different

• Define your own data model with entity templates, properties, thesauri and relationships, so records match your domain rather than a fixed schema
• Documents become searchable knowledge through OCR aware full text search over every uploaded file, powered by Elasticsearch
• Relate records to one another to map connections between people, events, cases and documents
• Publish a curated public library or keep the whole collection private, with granular control over what is shared
• Rich document viewer with in context references, table of contents and text selection

## Application Stack

• Uwazi 1.229 built from the official production release on Node.js 20
• nginx reverse proxy on ports 80 and 443 with a self signed certificate generated per instance
• MongoDB 7.0 as a single node replica set, published to the loopback interface only
• Elasticsearch 8.18 with the ICU analysis plugin for multilingual full text search, published to the loopback interface only
• Redis 6.0 backing the Uwazi job queue, published to the loopback interface only
• Reach the platform through nginx; the application, database, search and queue ports need no inbound rule and stay closed

## Secure By Default

Unlike images that ship a shared or documented password, this one carries no known credential:

• Uwazi's default administrator password is overwritten during the build, so the image ships with no usable login
• On first boot an administrator account is rotated to a per instance password written to a root only file
• A fresh session secret is generated on every first boot
• MongoDB and Elasticsearch are bound to the loopback interface, never exposed to the network
• No shared credentials exist anywhere in the image

## Storage Layout

Two dedicated EBS volumes ship with the image, separate from the operating system disk and each independently resizable: the Docker data root holding the MongoDB datastore and the Elasticsearch index, and the application tier holding the Uwazi bundle and your uploaded documents. Both are mounted by filesystem UUID.

## Getting Started

1. Launch the image on an m5.large instance or larger, with ports 80 and 443 open in your security group

2. Connect over SSH and read the generated administrator password from the root only credentials file

3. Browse to the instance address and sign in as the administrator

4. Create your first entity template in Settings, then upload a document and watch it become searchable

5. Relate records, build thesauri and decide what to publish to a public library

## Recommended Instance Sizing

• Minimum: m5.large (2 vCPU, 8 GB RAM) for small collections and evaluation
• Recommended: m5.xlarge (4 vCPU, 16 GB RAM) for active collections with heavy indexing
• Storage: grow the data volume as your document library grows
• Networking: ports 80 and 443 for browsers, port 22 for administration

## 24/7 cloudimg Support

Every deployment is backed by cloudimg engineers available around the clock by email and live chat. Support covers:

• Deployment guidance and instance sizing
• MongoDB and Elasticsearch administration, backup and restore
• nginx reverse proxy and TLS termination with Let's Encrypt or your own certificates
• Uwazi upgrades and migration assistance
• Performance tuning and troubleshooting

Critical issues receive a one hour average response time.

## Evaluation

Launch an instance to evaluate Uwazi with your team and terminate at any time with no long term commitment.

Uwazi is a trademark of HURIDOCS. All product and company names are trademarks or registered trademarks of their respective holders. Use of them does not imply any affiliation with or endorsement by them.