User Guide

The complete guide to every feature in the Operanix platform. Whether you are deploying your first agent or managing an enterprise AI workforce, this guide covers it all.

1. Welcome to Operanix

What Is Operanix?

Operanix is an Agent Operating System — a platform that lets businesses build, deploy, govern, and continuously improve a workforce of specialized AI agents. Unlike general-purpose chatbots, Operanix agents are trained on your company’s knowledge, evaluated for accuracy, governed by your compliance policies, and deployed through your customer-facing channels.

How Operanix Differs from ChatGPT

What Is an AI Agent?

An AI agent is a specialized virtual team member that handles tasks within a specific business domain. Each agent has its own identity (name, persona, voice), knowledge (your company data), and rules (what it can and cannot do). Think of an agent as a highly capable new hire that has read every document in your company and can answer questions 24/7.

What Is an AI Workforce?

An AI workforce is the full collection of agents in your organization. Operanix provides 11 pre-built agents covering every major department:

What Can You Accomplish?

• Customer Support — Deflect 60%+ of support tickets with accurate, instant AI responses trained on your help docs
• Sales Enablement — Instantly answer prospect questions about pricing, features, and integrations on your website
• Internal Knowledge — Give employees a single place to ask questions about policies, benefits, and procedures
• Content Creation — Generate on-brand marketing content, social posts, and email campaigns at scale
• Compliance & Risk — Automate contract review, policy checks, and regulatory compliance with governed AI
• Executive Intelligence — Get instant cross-department summaries and strategic briefings on demand

2. Platform Overview

The Operanix platform is organized into four main sections, accessible from the sidebar navigation.

3. Getting Started

The fastest path from signup to a deployed agent follows five steps. For a more detailed walkthrough, see the Quick Start Guide.

1. Sign up at app.operanix.ai and create your workspace.
2. Add knowledge — Crawl your website or upload documents in Knowledge Ops.
3. Review your agent — Install an agent from the Workforce, assign knowledge, and customize its identity.
4. Evaluate — Run an automated evaluation in Agent Readiness to verify quality.
5. Deploy — Push the agent to staging, test it, and then promote to production.

4. Your AI Workforce

Agent Lifecycle

Every agent follows a six-stage lifecycle from creation to continuous improvement:

• Build — Install or create an agent, configure its identity and persona
• Train — Assign knowledge sources so the agent can answer domain-specific questions
• Evaluate — Run automated evaluations to verify accuracy and readiness
• Deploy — Push to staging then production across your chosen channels
• Monitor — Track performance, trust scores, and conversation analytics
• Improve — Add knowledge, refine prompts, and re-evaluate based on real-world data

Configuring an Agent

Each agent has multiple configuration tabs:

Autonomy Levels

Certification Levels

5. Knowledge Operations

Knowledge is the foundation of agent quality. The better your knowledge base, the more accurate and helpful your agents will be.

Knowledge Categories

Adding Knowledge

There are two primary methods to add knowledge:

Website Crawl

1. Navigate to Knowledge Ops in the sidebar.
2. Click Add Source and select Website Crawl.
3. Enter your website URL (e.g., https://yourcompany.com).
4. Configure crawl depth and page limits (optional).
5. Click Start Crawl. Operanix will discover pages, extract content, and add them to your knowledge base.

File Upload

1. In Knowledge Ops, click Add Source and select File Upload.
2. Drag and drop files or click to browse. Supported formats: PDF, DOCX, XLSX, CSV, TXT, HTML, and Markdown.
3. Assign a category and optional tags.
4. Click Upload. Files are processed automatically (text extraction, chunking, embedding).

Agent Knowledge Assignment

After adding knowledge, assign it to agents:

1. Go to the agent’s Knowledge tab.
2. Click Assign Knowledge.
3. Select knowledge sources by category, individual document, or bulk select.
4. Save changes. The agent will immediately start using the assigned knowledge for responses.

Knowledge Matrix

The Knowledge Matrix view shows a grid of all agents vs. all knowledge sources, making it easy to see which agents have access to which knowledge and identify gaps in coverage.

6. Agent Readiness

Agent Readiness is your quality assurance system. It evaluates agents against real-world questions and scores them across multiple dimensions to determine deployment readiness.

Running an Evaluation

1. Navigate to Agent Readiness in the sidebar.
2. Select the agent you want to evaluate.
3. Click Run Evaluation. The system generates test cases from the agent’s knowledge base.
4. Wait for the evaluation to complete (typically 2–5 minutes).
5. Review results: pass rate, dimension scores, individual case results, and improvement recommendations.

Understanding Verdicts

Evaluation Dimensions

Each evaluation scores responses across seven quality dimensions:

1. Accuracy — Is the information factually correct?
2. Groundedness — Is the response based on knowledge base sources (not hallucinated)?
3. Helpfulness — Does the response actually help the user accomplish their goal?
4. Relevance — Does the response address the specific question asked?
5. Completeness — Does the response cover all aspects of the question?
6. Safety — Does the response avoid harmful, biased, or inappropriate content?
7. Citation Quality — Are claims properly attributed to specific knowledge sources?

Improvement Opportunities

After each evaluation, Operanix provides specific recommendations for improving agent quality:

• Knowledge gaps — topics where the agent lacks information
• Conflicting sources — cases where multiple knowledge sources provide contradictory information
• Prompt refinements — suggested changes to the agent’s system prompt
• Missing categories — knowledge categories that should be added

7. Deploying Agents

Deployment Channels

Widget Deployment

1. Go to the agent’s Deployment tab.
2. Select Chat Widget as the channel.
3. Customize the widget appearance: colors, position, welcome message, and logo.
4. Copy the embed code snippet.
5. Paste the snippet into your website’s HTML before the closing </body> tag.
6. The widget appears on your site immediately.

Widget Customization

• Brand colors — Match the widget to your brand palette
• Position — Bottom-right (default), bottom-left, or inline embed
• Welcome message — Custom greeting shown when the widget opens
• Avatar — Custom agent avatar or your company logo
• Lead capture — Collect visitor name and email before starting chat
• Suggested questions — Pre-defined questions shown as quick-start buttons

Hosted Chat

Each deployed agent automatically gets a hosted chat URL (e.g., chat.operanix.ai/your-company/sales). Share this link with anyone — no installation or embedding required. Hosted chat pages are mobile-responsive and support your custom branding.

Deployment Center

The Deployment Center provides a unified view of all deployed agents across all channels. From here you can:

• View deployment status for every agent and channel
• Promote agents from staging to production
• Rollback a deployment to a previous version
• Monitor real-time conversation activity

8. Governance

Governance is what makes Operanix enterprise-ready. It ensures your AI agents operate within the boundaries you define.

Governance Center

The Governance Center is your control panel for all governance settings. It displays your overall compliance posture, active policies, pending approvals, and trust score trends across your workforce.

Compliance Posture Levels

Trust Score

Every agent has a real-time trust score (0.00 – 1.00) calculated from five weighted dimensions:

Production Safety

Before an agent can be deployed to production, it must pass safety checks:

• Evaluation pass rate above the configured threshold
• Trust score above the auto-approve threshold
• All required governance policies assigned
• No unresolved compliance warnings

Policies

Policies are configurable rules that govern agent behavior. Types include:

• Content Safety — Block harmful, inappropriate, or off-brand content
• Topic Boundaries — Restrict agents to their domain (e.g., prevent a Support agent from giving legal advice)
• Compliance Rules — Enforce regulatory requirements (GDPR consent, HIPAA safeguards)
• Custom Policies — Define any rule using keyword matching, semantic detection, or regex patterns

Approval Queue

When an agent response falls below the auto-approve trust threshold or matches a policy requiring review, it is held in the Approval Queue. Reviewers can:

• Approve — Deliver the response as-is
• Edit & Approve — Modify the response before delivery
• Reject — Block the response and optionally provide a reason

Audit Trail

Every action on the platform is logged in an immutable audit trail. Access it from Governance Center → Audit Trail. Filter by date, user, agent, action type, or resource. Export audit logs for compliance reporting.

9. Departments

Departments let you organize your workspace by business function, providing scoped access control for larger teams.

Creating Departments

1. Go to Settings → Departments.
2. Click Create Department.
3. Enter a name and optional description.
4. Assign a department head (optional).

Assigning Agents and Members

• Agents: Assign from the agent’s settings page or from the department view.
• Members: Assign from Settings → Team by editing a member’s profile.

Scoped Access

Members with department-scoped roles (Manager, Operator) only see agents, knowledge, evaluations, and analytics within their assigned departments. This ensures teams cannot accidentally modify other departments’ resources.

10. Billing

Plan Comparison

Upgrading Your Plan

Go to Settings → Billing and click Upgrade. Select your new plan and confirm. Upgrades take effect immediately with prorated billing for the current period.

Usage Tracking

Monitor your usage from the Billing dashboard: active agents, knowledge pages, API calls, LLM tokens consumed, and estimated costs for the current period.

11. Security & SSO

SSO Providers

SSO Setup

1. Go to Settings → Security → SSO.
2. Select your provider and follow the configuration guide.
3. Enter Client ID, Client Secret, and Issuer URL (or upload SAML metadata).
4. Test the connection, then enable SSO.
5. Optionally enforce SSO-only login for all workspace users.

RBAC Roles

Operanix provides 10 roles for granular access control:

MFA

Enable multi-factor authentication from Profile → Security. Administrators can enforce MFA for all users from Settings → Security.

Data Security

• AES-256 encryption at rest
• TLS 1.3 encryption in transit
• Tenant-level data isolation
• Configurable data retention policies
• SOC 2 Type II, GDPR, and HIPAA compliance support

12. Workflows

What Are Workflows?

Workflows are automated sequences of actions triggered by events in the platform. They let you chain agent responses, governance checks, notifications, and integrations into repeatable processes without writing code.

Building a Workflow

1. Navigate to Workflows in the sidebar.
2. Click Create Workflow.
3. Select a trigger (e.g., new conversation started, evaluation completed, approval requested).
4. Add steps by dragging actions onto the canvas: agent response, condition check, send notification, call webhook, wait for approval.
5. Connect steps to define the flow sequence.
6. Test the workflow with sample data.
7. Activate the workflow. It will run automatically when the trigger fires.

Workflow Templates

Operanix includes pre-built workflow templates for common scenarios:

• Support Escalation — Route complex tickets from AI to human agents with full conversation context
• Lead Qualification — Qualify prospects through a series of agent questions and route to sales
• Content Approval — Generate marketing content, run compliance check, and queue for human approval
• Onboarding Flow — Guide new employees through HR questions and policy acknowledgments

13. Analytics & ROI

Executive Dashboard

The executive dashboard provides a high-level view of your AI workforce performance:

Cost Intelligence

The Cost Intelligence dashboard breaks down AI spending by agent, model, department, and time period. Use it to:

• Identify which agents drive the most cost
• Compare model costs (e.g., GPT-4 vs. Gemini)
• Track spending trends over time
• Set budget alerts and spending limits

Per-Agent Analytics

Each agent has its own analytics page showing conversation volume, response quality, trust score history, most-asked questions, knowledge usage, and customer satisfaction trends.

14. Marketplace

One-Click Install

The Agent Marketplace offers pre-built, ready-to-deploy agents for common use cases. Browse the marketplace, preview an agent’s capabilities, and click Install to add it to your workforce. Assign your knowledge and deploy — no configuration required.

Agent Bundles

15. Troubleshooting

My agent is giving wrong answers

This usually means the agent’s knowledge base is missing information or contains outdated content. Go to Knowledge Ops, search for the topic, and add or update the relevant knowledge. Then re-run an evaluation to verify the fix.

My agent has a 0% evaluation pass rate

A 0% pass rate typically means the agent has no knowledge assigned, or the assigned knowledge does not cover the evaluation questions. Go to the agent’s Knowledge tab and verify that knowledge sources are assigned. If they are, check that the knowledge content actually covers the topics being tested.

My agent’s trust score says “Need Data”

Trust scores require evaluation data to calculate. Run at least one evaluation from Agent Readiness to generate the initial trust score. The score will update automatically as more evaluations and conversations occur.

How do I add team members?

Go to Settings → Team and click Invite Member. Enter their email address and select a role. They will receive an email invitation to join your workspace. See the Admin Guide for detailed user management instructions.

How do I upgrade my plan?

Navigate to Settings → Billing and click Upgrade. Select your desired plan and confirm. Upgrades are effective immediately with prorated billing.

Do I need to write code?

No. Operanix is a fully no-code platform. You can create agents, add knowledge, run evaluations, configure governance, and deploy — all through the visual interface. The only code required is the embed snippet if you choose to deploy the chat widget on your website.

How secure is my data?

Operanix uses AES-256 encryption at rest, TLS 1.3 in transit, and strict tenant-level data isolation. Your data is never shared with other organizations or used to train AI models. See the Security & Compliance page for full details.

Can agents access my internal systems?

Agents only access the knowledge you explicitly provide through Knowledge Ops (crawls, uploads, manual entries). They do not have access to your internal databases, file systems, or third-party tools unless you configure specific integrations through the Workflows or API.