Section 1: Executive Summary & Product Overview

1.1 What is Quantra?

Quantra is a general-purpose, visual, no-code pipeline builder designed for document processing and data orchestration. It empowers organizations across every industry to design, execute, and manage sophisticated document workflows without writing a single line of code. Through an intuitive drag-and-drop canvas interface, users construct pipelines that ingest data from diverse sources, apply powerful AI and machine learning transformations, and deliver processed results to interactive workbenches for human review, redaction, and decision-making.

At its core, Quantra addresses a fundamental challenge that organizations of all sizes face: the need to process, analyze, classify, and act upon large volumes of documents and unstructured data. Whether the task involves extracting text from scanned images, detecting and redacting personally identifiable information, classifying documents by type, summarizing lengthy reports, or deduplicating massive file repositories, Quantra provides a unified platform that brings all of these capabilities together under a single, coherent workflow engine.

Quantra is not tied to any specific industry or vertical. It is architected as a horizontal platform that can be applied to legal discovery, regulatory compliance, data migration, content management, digital archival, research data processing, and countless other domains where document-centric workflows are essential. Its extensible plugin architecture means that new data sources, processing tools, and workbench experiences can be added without modifying the core platform, ensuring that Quantra can adapt to the unique requirements of any organization or use case.

The platform is built on a distributed microservice architecture that communicates over gRPC with mutual TLS (mTLS) encryption, ensuring both high performance and enterprise-grade security. Pipelines execute in a streaming fashion, providing real-time progress feedback to users as documents flow through each stage of processing. The result is a system that is as responsive and transparent as it is powerful.

1.2 Key Capabilities

Visual No-Code Pipeline Design

Quantra's canvas UI is the primary interface through which users design document processing pipelines. The canvas provides a visual, node-based editor where each node represents a discrete step in the pipeline: a data source, a processing tool, or a workbench destination. Users drag nodes onto the canvas, configure their parameters through intuitive property panels, and connect them with edges that define the flow of documents through the pipeline. There is no programming required. The entire pipeline design process is visual, interactive, and immediate.

The canvas supports complex pipeline topologies including linear chains, branching paths, merging flows, and multi-stage processing graphs. Users can preview pipeline configurations before execution, save and load pipeline designs, and share them across teams. The visual nature of the canvas makes it easy to understand, audit, and modify even the most complex workflows at a glance.

Data Source Connectivity

Quantra connects to a broad range of data sources out of the box, enabling organizations to ingest documents from wherever they reside. Supported data sources include local and network file shares (including SMB/CIFS and NFS), relational databases, cloud storage services (such as Amazon S3, Azure Blob Storage, and Google Cloud Storage), email servers (via IMAP and Exchange protocols), and more. Each data source is implemented as a plugin, meaning that new source types can be added to the platform without altering the core system.

Data source nodes in the pipeline handle authentication, connection management, file enumeration, and content retrieval. They support filtering by file type, date range, folder path, and other metadata criteria, allowing users to precisely target the documents they wish to process. Data sources emit documents into the pipeline stream, where they are picked up by downstream processing tools.

AI/ML Document Processing

The heart of Quantra's value lies in its rich library of AI and machine learning processing tools. These tools operate on documents as they flow through the pipeline, extracting information, detecting patterns, transforming content, and enriching metadata. The platform ships with a comprehensive set of built-in tools, including:

• Optical Character Recognition (OCR): Extracts machine-readable text from scanned documents, images, and PDFs using state-of-the-art OCR engines. Supports multiple languages, handwriting recognition, and layout-aware text extraction.
• PII Detection: Identifies personally identifiable information within document text, including names, addresses, social security numbers, email addresses, phone numbers, financial account numbers, and other sensitive data patterns. Supports configurable sensitivity levels and custom pattern definitions.
• Document Classification: Automatically categorizes documents by type, topic, or content using machine learning models. Supports both pre-trained classifiers and custom models trained on organization-specific document taxonomies.
• Summarization: Generates concise summaries of lengthy documents using natural language processing, enabling rapid review and triage of large document sets.
• Face Detection: Identifies and locates human faces within images and document scans, supporting privacy workflows that require face redaction or anonymization.
• Fingerprint Analysis: Detects fingerprint images within documents, useful for forensic and identity-related document processing workflows.
• Hashing: Computes cryptographic hashes of documents and their contents for integrity verification, chain-of-custody tracking, and deduplication purposes.
• Deduplication: Identifies and flags duplicate or near-duplicate documents across large collections, reducing storage costs and review burden by eliminating redundant content.

Each processing tool is implemented as a plugin and runs as an independent microservice, communicating with the core platform over gRPC. This architecture ensures that tools can be developed, deployed, updated, and scaled independently of each other and of the platform itself.

Interactive Workbenches

After documents have been processed through a pipeline, they arrive at workbench nodes where human reviewers can interact with the results. Workbenches provide rich, purpose-built user interfaces for reviewing, annotating, redacting, approving, and exporting processed documents. The platform includes workbenches tailored for different review tasks, such as document redaction workbenches that allow reviewers to visually select and redact sensitive content, classification review workbenches that present AI-suggested categories for human confirmation, and general-purpose document viewers for browsing and inspecting pipeline output.

Workbenches are also implemented as plugins, meaning that organizations can develop custom workbench experiences that match their specific review workflows and user interface requirements.

Enterprise Security

Quantra is built with enterprise security requirements at its foundation. The platform supports multi-factor authentication (MFA) for user login, ensuring that access to sensitive document processing workflows is protected by strong identity verification. All inter-service communication is secured with mutual TLS (mTLS), providing both encryption in transit and mutual authentication between the platform and its microservices. Comprehensive audit logging captures every significant action taken within the platform, including pipeline executions, document access, user logins, configuration changes, and administrative operations, providing a complete audit trail for compliance and forensic purposes.

Extensible Plugin Architecture

Every major extension point in Quantra is exposed through a well-defined plugin interface. Data sources, processing tools, and workbenches are all plugins that can be added, removed, and updated independently of the core platform. This architecture ensures that Quantra can grow with an organization's needs, accommodating new document types, new processing techniques, new data sources, and new review workflows without requiring changes to the platform itself. Plugins are versioned independently and can be developed by third parties or by the organization's own development teams.

Distributed Microservice Architecture

Quantra's processing tools and workbenches run as independent microservices that communicate with the core platform via gRPC, a high-performance remote procedure call framework. This distributed architecture provides several key benefits: individual services can be scaled horizontally to handle increased load; services can be deployed across multiple machines or containers for fault tolerance; and services can be updated or replaced without downtime to the rest of the platform. The use of gRPC ensures low-latency, high-throughput communication between services, while mTLS secures every connection.

Streaming Execution with Real-Time Progress

Quantra executes pipelines in a streaming fashion, meaning that documents begin flowing through downstream processing stages as soon as they are emitted by upstream stages, rather than waiting for an entire batch to complete before proceeding. This streaming execution model minimizes latency, maximizes throughput, and provides users with real-time visibility into pipeline progress. As documents are processed, the UI updates in real time to show how many documents have been ingested, how many are currently being processed at each stage, and how many have completed, giving users confidence that their pipelines are running correctly and providing accurate time-to-completion estimates.

1.3 Use Cases

Because Quantra is a general-purpose platform, it can be applied to a wide variety of document processing and data orchestration scenarios across every industry. The following use cases illustrate the breadth of problems that Quantra can solve.

Legal Discovery (eDiscovery)

Law firms and corporate legal departments use Quantra to process large volumes of documents during litigation discovery. Pipelines ingest documents from file shares, email archives, and cloud storage; apply OCR to scanned materials; detect and flag privileged or confidential content using PII detection and classification tools; deduplicate the collection to reduce review volume; and deliver the processed set to review workbenches where attorneys can examine, tag, and redact documents before production. The visual pipeline designer makes it easy for litigation support professionals to configure and adjust discovery workflows without relying on IT or development resources.

Regulatory Compliance

Organizations in regulated industries use Quantra to ensure that their document repositories comply with data protection regulations such as GDPR, CCPA, HIPAA, and others. Pipelines scan document stores for personally identifiable information, classify documents by sensitivity level, generate compliance reports, and route documents requiring remediation to workbenches where compliance officers can review and redact sensitive content. The platform's audit logging capabilities provide the evidence trail that regulators require.

Data Migration

When organizations migrate from legacy document management systems to modern platforms, Quantra serves as the processing layer that transforms, enriches, and validates documents during the migration. Pipelines extract documents from legacy sources, apply OCR to digitize paper-based records, classify documents to apply new taxonomy labels, hash documents for integrity verification, and deduplicate the collection to avoid migrating redundant content. The visual pipeline designer allows migration project managers to design and iterate on migration workflows rapidly.

Content Processing and Publishing

Media companies, publishing houses, and content-driven organizations use Quantra to process incoming content at scale. Pipelines ingest manuscripts, articles, images, and multimedia files from submission portals and email; apply OCR and text extraction; classify content by topic, genre, or format; generate summaries for editorial review; and detect faces in images for privacy compliance. Processed content is delivered to editorial workbenches where reviewers can approve, annotate, and prepare materials for publication.

Digital Archival and Records Management

Libraries, government agencies, museums, and other institutions responsible for preserving large document collections use Quantra to digitize, catalog, and manage their archives. Pipelines ingest scanned documents and images, apply OCR to make content searchable, classify materials by type and era, compute hashes for integrity and provenance tracking, detect and redact sensitive information in public records, and deduplicate collections that have accumulated redundant copies over decades. Workbenches allow archivists to review, annotate, and curate processed materials before they are ingested into long-term preservation systems.

Financial Document Processing

Banks, insurance companies, and financial services firms use Quantra to process the high volumes of documents that accompany lending, underwriting, claims processing, and regulatory reporting. Pipelines extract data from loan applications, insurance claims, and financial statements using OCR; classify documents by type (pay stubs, tax returns, bank statements, etc.); detect PII for privacy compliance; and route processed documents to review workbenches where underwriters and claims adjusters can make decisions. The streaming execution model ensures that time-sensitive documents are processed with minimal latency.

Human Resources and Recruitment

HR departments use Quantra to process resumes, employment applications, onboarding documents, and personnel files. Pipelines ingest documents from email and applicant tracking systems, extract text with OCR, classify documents by type, detect and redact PII for anonymized review processes, and deliver processed documents to HR workbenches for review and decision-making. The no-code pipeline designer allows HR operations staff to configure and adjust workflows without technical assistance.

Research and Academic Data Processing

Research institutions and universities use Quantra to process large collections of academic papers, research data, grant applications, and institutional records. Pipelines classify documents by research domain, extract and summarize key findings, detect duplicate submissions, and prepare materials for repository ingestion. The extensible plugin architecture allows research teams to integrate domain-specific processing tools tailored to their particular fields of study.

1.4 Target Audience

Quantra is designed for a diverse range of users and roles within an organization. Its no-code visual interface makes it accessible to non-technical business users, while its extensible architecture and microservice foundation make it a powerful tool for technical teams as well.

Business Users and Process Owners

Business analysts, compliance officers, litigation support professionals, records managers, and other non-technical users who own document processing workflows are the primary users of Quantra's canvas UI. These users design pipelines, configure processing parameters, execute workflows, and review results through workbenches, all without needing programming skills or IT support. Quantra empowers these users to take direct control of their document processing needs, reducing dependency on technical teams and accelerating time to value.

IT Administrators and DevOps Engineers

IT professionals responsible for deploying, configuring, securing, and maintaining the Quantra platform are a key audience for this documentation. These users manage the platform's installation, configure data source connections, set up mTLS certificates, manage user accounts and MFA policies, monitor microservice health, and ensure that the platform operates reliably and securely within the organization's infrastructure.

Developers and Integration Engineers

Software developers who extend Quantra by building custom plugins (data sources, processing tools, and workbenches) or who integrate Quantra with other systems in the organization's technology stack. These users work with Quantra's plugin APIs, gRPC service definitions, and extension points to build new capabilities that address organization-specific requirements.

Executives and Decision Makers

Leaders evaluating Quantra as a platform investment will find the executive summary and product overview sections of this documentation useful for understanding the platform's capabilities, architecture, and value proposition. Quantra's ability to reduce manual document processing effort, improve compliance posture, accelerate discovery timelines, and provide audit-ready processing trails are key value drivers for executive stakeholders.

1.5 Product Highlights

1.6 How It Works

At a high level, Quantra operates through a straightforward workflow that takes users from pipeline design to document review in a series of intuitive steps. The following overview describes the end-to-end process.

Step 1: Design the Pipeline

The user opens Quantra's canvas UI and begins designing a document processing pipeline. They drag data source nodes onto the canvas to define where documents will be ingested from: a network file share, a cloud storage bucket, an email mailbox, or any other supported source. They configure each data source with connection credentials, folder paths, file type filters, and other parameters.

Next, the user adds processing tool nodes to the canvas and connects them to the data source nodes with edges that define the document flow. Each tool node is configured with its specific parameters: the OCR engine and language settings for an OCR node, the sensitivity level and pattern rules for a PII detection node, the model and category set for a classification node, and so on. Users can chain multiple tools together in sequence, branch the pipeline to apply different tools in parallel, or merge multiple streams back together for downstream processing.

Finally, the user adds workbench nodes at the end of the pipeline to define where processed documents will be delivered for human review. Workbench nodes are configured with review parameters such as the workbench type, reviewer assignments, and display settings.

Step 2: Execute the Pipeline

When the pipeline design is complete, the user initiates execution with a single click. Quantra's execution engine takes over, orchestrating the flow of documents through the pipeline. The platform connects to the configured data sources and begins enumerating and retrieving documents. As each document is retrieved, it is streamed to the first processing tool in the pipeline, which performs its analysis or transformation and passes the result to the next tool in the chain.

The streaming execution model means that documents do not wait for the entire collection to be ingested before processing begins. As soon as the first document is available, it starts flowing through the pipeline. This approach minimizes end-to-end latency and ensures that reviewers can begin their work as soon as the first processed documents arrive at the workbench, even while the pipeline continues processing the remaining documents in the background.

Throughout execution, the canvas UI displays real-time progress indicators on each node, showing the number of documents that have entered, are currently being processed by, and have exited each stage. Error counts and status indicators provide immediate visibility into any issues that arise during processing.

Step 3: Review and Act

As processed documents arrive at workbench nodes, they become available for human review. Reviewers open the workbench interface, which presents processed documents along with all of the metadata, annotations, classifications, and extracted data that the pipeline's tools have produced. Depending on the workbench type, reviewers can perform a variety of actions: visually redacting sensitive content that PII detection has flagged, confirming or correcting AI-suggested document classifications, reviewing and approving summarized content, examining deduplicated document clusters, and more.

Workbench actions are recorded in the platform's audit log, providing a complete record of every review decision for compliance and quality assurance purposes. Once review is complete, documents can be exported, archived, or routed to downstream systems as needed.

Step 4: Iterate and Refine

Quantra's visual pipeline designer makes it easy to iterate on pipeline designs based on review results. If reviewers find that the OCR tool is not extracting text accurately from a particular document type, the user can adjust the OCR configuration and re-run the pipeline. If the PII detection tool is generating too many false positives, sensitivity thresholds can be tuned. If a new processing step is needed, a new tool node can be added to the pipeline without rebuilding the entire workflow from scratch. This iterative approach allows organizations to continuously improve their document processing pipelines over time, adapting to new requirements, new document types, and lessons learned from review.

Under the Hood

Behind the visual interface, Quantra operates as a Django-based web application backed by a distributed constellation of microservices. The core platform manages pipeline definitions, user authentication, job scheduling, and the canvas UI. Processing tools and workbenches run as independent microservices, each listening on its own port and communicating with the platform over gRPC with mTLS encryption. When a pipeline is executed, the platform's orchestration engine coordinates the microservices, routing documents between them according to the pipeline's topology, managing buffering and backpressure, and aggregating progress information for the UI.

This architecture ensures that the platform remains responsive and stable even under heavy load, that individual tools can be updated or scaled without affecting the rest of the system, and that the entire processing infrastructure is secured with modern cryptographic protocols. The result is a platform that is as robust and secure under the hood as it is intuitive and accessible on the surface.