Synetic Admin

Dashboard

Platform overview · Loading…

Organizations

Agents

End Users

Messages · 30d

Quick actions

Recent organizations

View all →

Name	Owner	Status	Created

Costs

Billing

Organization subscriptions, revenue, and billing status

Stripe Dashboard

Organizations

Customer accounts, agents, and platform broadcasts

Broadcasts 0

Loading…

All Organizations

ID	Name	Email	Agents	Status	Last Login
Loading…

Organization Detail

Loading…

Team

Manage who can access the Synetic admin panel

Platform Admins

Name (optional)

Name	Email	Status	Last Login

Organization Plans

Configure agent plans, features, and settings per organization type

Loading…

Agent Plans

Configure plans and their included features

Create New Plan

Key

Lowercase, alphanumeric, e.g. starter

Name

Monthly Price ($)

Sort Order

Storage Cap (MB)

Description

Loading plans…

Feature Library

Platform and marketing configuration for each feature

Saved ✓

Loading…

Platform Integrations

Authentication, database, storage, and third-party service configuration

Platform integrations are server-level services that the API depends on. Changes to credentials update the .env file and require a service restart to take effect.

Loading…

Update Integration

System

Platform configuration and health

Feature Flags

new_editor

Next-gen agent editor interface

billing_portal

Self-serve billing management

rag_v2

Retrieval-augmented generation v2 pipeline

Platform Settings

max_tenants_per_account

max_agents_free

max_agents_pro

trial_days

Database Health

PostgreSQL

Connected

Checking version…

Settings

Admin profile and access control

Your Profile

—

Synetic Admin Emails

The following emails have super-admin access, managed via the SYENTIC_ADMIN_EMAILS environment variable.

Loading…

Guardrails

Manage content safety guardrails for all AI agents

Mandatory — always active for all agents, cannot be disabled by organization owners.
Optional — available to agents whose plan includes the guardrails feature. Organization owners toggle per agent.
Rail Type — input checks user messages, output checks AI responses, both checks both.

Loading…

New Guardrail

Key (unique)

Name

Rule — for judge-based rails this text is what runs at every turn

Category

Rail Type

Order

Mandatory Active

Config (JSON) — runtime-consumed; leave empty for judge-only rails

Schema by rail type:
input_redact / output_redact: { "patterns": [{ "label": "ssn", "regex": "...", "replacement": "[SSN]" }] }
output_warn deterministic (e.g. hallucination):

{ "deterministic": { "signal": "knowledge_fidelity", "operator": "<", "threshold": 60, "requires": ["rag_active"] }, "warn_message": "..." }

How the rule runs: For input_block, output_block, and output_warn, the runtime LLM judge consumes the Rule text above on every applicable turn — that text is what enforces. For input_redact and output_redact the patterns now come from the Config (JSON) field — admins can extend the pattern set without a code change. For deterministic output_warn rails (e.g. hallucination), the threshold and warning message also live in Config (JSON). (NeMo Colang is no longer used; the legacy colang and config_yaml columns remain in the schema for back-compat but are not executed.)

Release Notes

Publish changelog entries visible to organization owners

Version	Title	Status	Date

Marketing CMS

Manage marketing site text content

Loading…

Monitoring

Service health and platform metrics

Data & Storage Utilization

Loading…

Organization Storage

Organization	Plan	Users	DB Storage

Recent Events (1h)

Loading…

Logs

Track all admin actions

Audit log

Time	Actor	Action	Target	Detail

Audit Log

Cross-org event history with cohort, subscriber, and success scoping.

Events

When	Actor	Action	Target	Org	Success

Expert System

Architecture, evaluation, safety, and privacy reference for the Synetic Expert Agent platform. Read-only — for canonical text see Information → Research docs.

Open in new tab

Information

Platform features catalog and live documentation from the monorepo. All Manage app users can browse here.

Documentation

Loading…

On this page