Scalix Prime User Guide

What is Scalix Prime?

Scalix Prime is an enterprise data platform that connects all your information into a knowledge graph — a living map of how your data relates. Instead of keeping data in isolated spreadsheets, databases, and systems, Prime brings it together so you can see the full picture.

With Scalix Prime, you can:

• Explore your data visually — Browse entities and their relationships across departments and systems.
• Ask questions in plain English — AI agents search your data and return answers grounded in real information.
• Build apps without coding — Create dashboards, operational tools, and reports using drag-and-drop widgets.
• Automate workflows — Design multi-step AI pipelines that search, reason, and take action on your data.
• Transform and analyze data — Clean, filter, aggregate, and visualize data with point-and-click tools.

Everything runs through the Workshop — a browser-based workspace where you access all of these tools from a single sidebar.

---

Key Concepts

Domains

A domain is a top-level grouping for related data — like healthcare, financial-services, or supply-chain. Each domain has its own entity types, relationships, and data schema. When you select a domain in the Workshop, you're scoping everything to that data boundary.

Entity Types & Entities

An entity type is a category of objects — Patient, Account, Supplier. Each entity is a specific instance with properties (like a patient's name, date of birth, and medical record number) and relationships to other entities.

Relationships

Relationships are connections between entities. For example, a Patient HAS_CONDITION Condition, or an Account TRANSFERS_TO Account. These connections are what make a knowledge graph powerful — they reveal patterns that isolated data can't show.

Ontology

The ontology is the blueprint of your knowledge graph. It defines which entity types exist, what properties they have, and how they connect. Scalix Prime observes your data patterns and recommends ontology improvements for your approval.

---

Quick Start

Get started with Scalix Prime Workshop in three steps:

1. Open the Workshop — Navigate to your Workshop URL. You'll land on the Home page with quick access to all tools and your saved apps.
2. Explore your data — Go to Ontology Explorer, select a domain, and browse entity types. Click any entity to see its properties and relationships.
3. Ask a question — Open AI Chat, pick an agent, and ask a question about your data in plain English. The agent will search the knowledge graph and return results.

---

Ontology Explorer

The Ontology Explorer is a three-panel browser for navigating your knowledge graph. Select a domain, pick an entity type, and drill into individual entities and their relationships.

How to use it

1. Select a domain — Use the dropdown in the left panel to choose which data area to explore (e.g., Healthcare, Financial Services).
2. Choose an entity type — Click a type button like Patient or Account. The center panel loads all entities of that type.
3. Browse entities — Search or scroll through the list. Click any row to open its detail panel on the right.
4. Explore relationships — In the detail panel, expand relationship groups to see connected entities. Click a related entity to navigate to it.

Use the Open in Graph button to see an entity and its neighborhood as an interactive visual diagram.

---

AI Chat

AI Chat lets you interact with your knowledge graph using natural language. Ask questions, run searches, and trigger actions through specialized AI agents.

Choosing an agent

• General — For open-ended questions about your data and platform capabilities.
• Data Query — Specialized for searching the knowledge graph. Translates your questions into structured queries.
• Action — Executes workflows: creates entities, triggers pipelines, or updates data. Always asks for confirmation first.
• Admin — Manages ontology schemas, monitors system health, and configures settings.
• Analytics — Computes aggregations, spots patterns, and generates data summaries.

How it works

Type your question in plain English. The agent will search the knowledge graph, call relevant tools, and return an answer grounded in your actual data. You'll see tool calls as expandable sections in the conversation — click them to see exactly what the agent searched for and what it found.

---

App Builder

Build data applications without writing code. Create dashboards, operational tools, and analytical views by dragging widgets onto a grid canvas.

How to use it

1. Create or open an app — From the Home page, click New App or open an existing one.
2. Add widgets — Drag widgets from the left palette onto the canvas. Options include charts, tables, search bars, graph viewers, KPI cards, forms, timelines, maps, and more.
3. Configure — Click any widget to open its settings on the right. Set the data source (domain + entity type), display options, column mappings, and event handlers.
4. Preview & share — Click Preview to see your app as end users will experience it. Use Export to download the app definition, or Share to copy a preview URL.

Widget types

---

Agent Studio

Build multi-step AI workflows visually. Chain together search, reasoning, transformation, and action steps to automate complex data tasks.

Node types

• Input — Entry point. Receives a user query, webhook payload, or scheduled trigger.
• Search — Searches the knowledge graph using text, semantic, or hybrid search.
• Reason — Sends data to an AI model for analysis. Adjust the temperature slider for focused (0) or creative (2) responses.
• Transform — Manipulates data — extract fields, format text, compute values.
• Condition — Branches the flow based on a true/false expression.
• Action — Executes operations — create entities, send notifications, call external APIs.
• Output — Returns the final result.

Building a pipeline

1. Add nodes — Drag nodes from the left palette onto the canvas.
2. Connect them — Drag from a node's output handle to another node's input handle to create a data flow.
3. Configure each node — Click a node to set its parameters in the right panel.
4. Run the pipeline — Click Run Pipeline. Nodes light up green as they complete or red on failure.

---

Data Pipeline

Import, clean, transform, and load data — all visually. Bring data in from files or APIs, process it step by step, and load it into the knowledge graph or export it.

Step types

• Source — Where data comes from: CSV upload, PostgreSQL, REST API, S3 bucket, or an existing domain entity.
• Clean — Remove nulls, trim whitespace, normalize case, remove duplicates, fix date formats.
• Transform — Rename columns, cast types, compute new fields, split or merge columns.
• Filter — Keep only rows matching a condition (equals, contains, greater than, regex, etc.).
• Join — Combine two data sources on a matching key. Supports inner, left, right, and full outer joins.
• Aggregate — Group by field(s) and compute count, sum, average, min, or max per group.
• Destination — Where results go: create ontology entities, export CSV, or POST to a webhook.

---

Analysis Workbench

Explore and visualize data without writing queries. Load data from any domain, apply transformations, and view results as tables or charts.

Workflow

1. Select data — Choose a domain and entity type from the dropdowns, then click Load Data.
2. Build a pipeline — Add analysis steps: Filter, Sort, Group, Aggregate, Column Select, or Limit. Each step transforms the data for the next.
3. Visualize — Switch between Table, Chart, or Split view. In chart mode, pick a chart type (bar, line, pie, scatter) and map your X and Y axes.
4. Export — Download your results as CSV for use in Excel or other tools.

---

Administration

The Admin panel provides tools for managing your platform. It is organized into tabs:

• Ontology — Browse and manage your domains, entity types, relationships, and properties.
• Monitoring — View system health, component status, and uptime metrics.
• Data Quality — See quality scores across dimensions like completeness, consistency, and accuracy. Review and resolve flagged issues.
• Standards — Browse industry standards (FHIR, MITRE ATT&CK, STIX, Schema.org) and import them into your domains.
• Classification — Configure security classification levels and redaction rules for sensitive data.
• Extensions — View loaded engine extensions and their status.
• MCP Tools — Browse the tools available to the AI system for queries, mutations, and admin tasks.
• Entity Resolution — Find and merge potential duplicate entities using fuzzy matching.

---

Code Intelligence

Map your entire codebase as an interactive knowledge graph. Scalix Prime parses source code in 6 languages (Python, TypeScript, Go, Rust, Java, C#) and builds a graph of modules, functions, classes, dependencies, and call chains.

What you can do

---

Healthcare & Life Sciences

Scalix Prime includes built-in support for the HL7 FHIR R4 healthcare standard, mapping patient records, conditions, medications, and encounters into a standards-compliant knowledge graph.

What you can do

---

Supply Chain & Logistics

Model your entire supply chain as a connected graph — every supplier, sub-supplier, warehouse, shipment route, and material flow. See the ripple effects of disruptions before they hit your production line.

What you can do

---

Defense & Intelligence

Built for environments where security is non-negotiable. Scalix Prime runs completely offline — no internet, no cloud dependency, no telemetry. Classification labels are enforced on every piece of data.

What you can do

---

Financial Services

Trace transaction flows across accounts, institutions, and jurisdictions. Scalix Prime's real-time graph queries make it viable for fraud detection, AML/KYC compliance, and risk monitoring.

What you can do

---

Platform Capabilities

Scalix Prime combines several technologies into one integrated platform:

---

API Reference

Every operation available in the Workshop is also available through the REST API. Enterprise clients can integrate directly without using the drag-and-drop UI — ingest data, query the graph, manage schemas, and run AI agents all via HTTP.

Health & Status

Data Ingestion

Create entities and relationships in the knowledge graph programmatically. Supports single, batch, and CSV upload.

Search & Traversal

AI Agents & Reasoning

Domain & Schema Management

Training & Drift Detection

Connectors

Code Analysis

Example: Ingest and query via curl

# Create an entity
curl -X POST https://your-instance/api/v1/ingest/entity \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "domain": "finance",
    "entity_type": "Transaction",
    "name": "TXN-2024-001",
    "properties": {"amount": 50000, "currency": "USD"}
  }'

# Search the knowledge graph
curl -X POST https://your-instance/api/v1/search \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"query": "high risk transactions over 10000", "limit": 20}'

# Ask a question via the agent pipeline
curl -X POST https://your-instance/api/v1/query \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"query": "Which accounts have circular fund flows?", "domain": "finance"}'

---

Performance

Scalix Prime is built for speed. Here's what that means for your day-to-day experience:

• Real-time graph queries — Navigate complex data relationships in under a millisecond. Finding how two entities are connected across 3 hops takes less than 1μs.
• Instant search — Combined text, semantic, and fuzzy search returns results in milliseconds across your entire knowledge graph.
• Live permission checks — Access control is enforced on every single property access without any noticeable delay.
• No waiting for queries — Whether you're exploring data in Ontology Explorer, running analysis, or asking the AI a question, results come back fast enough for real-time interaction.

---

Security & Compliance

Your data is isolated

Every customer gets their own dedicated database instance. Your data is physically separated from other customers' data — not just logically partitioned. Cross-tenant data access is structurally prevented.

Encryption & integrity

• In transit — All network communication uses TLS 1.3.
• At rest — Customer-managed encryption keys (CMEK) and automated key rotation on the roadmap.
• Data integrity — WAL protected with CRC32 integrity checksums. Every write is verified on read.

Fine-grained access control

Access permissions go all the way down to individual properties on individual entities. For example, a nurse can see a patient's vitals but not their psychiatric notes. Roles inherit permissions from parent roles, and wildcard rules make broad grants easy while keeping fine-grained overrides simple.

Compliance

---

Smart Schema

Scalix Prime includes a schema evolution system that improves your data model over time. Instead of manually defining every entity type and relationship, the platform observes how your data behaves and recommends changes for your approval.

How it works

1. Observe — The system monitors data operations: entity creation, property updates, relationship changes, query patterns, and access frequencies.
2. Recommend — Based on observed patterns, it suggests new entity types, missing relationships, property additions, index optimizations, and schema refinements.
3. You decide — Every recommendation requires your review and approval before it's applied. The system never changes your data model without permission.
4. Drift alerts — If new data patterns diverge from your current schema, you'll be alerted before applications break — especially valuable when data sources change without notice.

Built-in industry standards

You can import pre-built ontology schemas for common industry standards:

• HL7 FHIR R4 — Healthcare interoperability (Patient, Condition, Observation, etc.)
• MITRE ATT&CK v14 — Cybersecurity threat taxonomy
• STIX 2.1 — Structured Threat Intelligence eXpression
• Schema.org — General-purpose linked data vocabulary

Import any of these from the Administration > Standards tab in Workshop.

---

Frequently Asked Questions

Do I need to know SQL or Cypher?

No. The AI Chat agents translate your plain English questions into queries automatically. Workshop tools like Ontology Explorer and Analysis Workbench are fully point-and-click. If you do know SQL or Cypher, the Query tool in Chat lets you run them directly.

Can I export my data?

Yes. The Analysis Workbench lets you download results as CSV. App Builder can export app definitions as JSON. Data Pipeline supports exporting to CSV and webhooks.

What happens if the AI gives a wrong answer?

AI Chat agents always ground their answers in your actual data. You can see exactly what the agent searched for and what it received by expanding the tool call sections in the conversation. The data shown is real — the AI interprets it but doesn't fabricate data points.

Can multiple people use the Workshop at the same time?

Yes. Each user has their own session, saved apps, and conversation history. Access control ensures users only see data they're authorized to access.

How do I add my own data?

Use the Data Pipeline tool to import data from CSV files, databases, APIs, or S3 buckets. The pipeline lets you clean and transform the data before loading it into the knowledge graph.

How do I get help inside the Workshop?

Every page has a Tour button that walks you through the key features. Look for the small ? icons next to complex controls for contextual explanations. You can also ask the AI Chat General agent questions about how to use the platform.

Can I use Scalix Prime without the Workshop UI?

Yes. Every operation available in the Workshop is exposed through the REST API at /api/v1/. You can ingest data, search the graph, manage schemas, run AI agents, and administer connectors entirely via curl, Python, or any HTTP client. Interactive API docs are served at /docs (Swagger) and /redoc. See the API Reference section for details.

Can Scalix Prime run without internet?

Yes. Scalix Prime supports fully air-gapped deployment for defense and high-security environments. The engine runs as a self-contained binary with no external dependencies.