WeGen-AIWeGen-AI
Documentation
Documentation

Welcome to WeOS

The AI operating system for enterprises. Everything you need to build, orchestrate and govern AI agents at enterprise scale.

What is WeOS?

WeOS is an AI operational platform designed for enterprises that need control, governance and scale. It is the infrastructure upon which your AI agents operate.

The operating system analogy: Just as Windows or Linux manage processes, memory and devices of a computer, WeOS manages agents, connections, decision flows and your company's data. AIWares are the "software" that runs on top of this infrastructure.

Two layers, one operation

WeOS

WeOS — The Infrastructure

Orchestrates agents, manages connections, audits decisions, controls access and ensures compliance. The "house" where everything runs.

AIWare

AIWare — The Software

Specific configuration of agents, flows and rules for a business process. Runs on WeOS. It is your product, with your brand.

Key Capabilities

Agent Orchestration

Centralized management of all agents with persistent memory and enterprise context.

Architecture-First Governance

Granular RBAC, immutable audit trail and LGPD compliance built-in from design.

Universal Integration

+200 native connectors for CRM, ERP, databases, APIs and communication channels.

Total Observability

Per-agent costs, decision logs, flow performance and real-time alerts.

Platform Resources

WeOS is composed of six main modules that work together to create a complete enterprise AI platform.

AIWare

AIWares are vertical AI solutions configured on top of WeOS. Each AIWare groups agents, flows and knowledge bases to solve a specific business process.

AIWare has no standalone API: All access to an AIWare is done through the WeOS API. AIWare is a configuration within the platform, not an independent service. This ensures centralized governance of all operations.

AIWare Catalog

Autonomous SDR

24/7 lead qualification with profile analysis and CRM updates

CS with Real Memory

Omnichannel support with real memory and knowledge base access

HR Screening & Onboarding

From CV screening to personalized employee onboarding

Financial Reconciliation

Autonomous bank reconciliation with anomaly detection

Contract Analysis

Semantic search and risk analysis in legal repositories

Custom AIWare

Built for your specific process, with your data and rules

Agentes

Agentes são a força de trabalho digital do WeOS. Executam tarefas, tomam decisões e aprendem com o contexto da sua empresa. Possuem memória persistente de curto e longo prazo.

Memória Persistente

Curto prazo (sessão) e longo prazo (workspace)

Multi-modelo

GPT-4o, Claude, Gemini, Llama e mais de 20 modelos

Especialização

Configurados por departamento e processo

Conexões

Integre o WeOS com qualquer sistema em minutos. +200 conectores nativos para CRM, ERP, banco de dados, APIs, planilhas e canais de comunicação. Seus sistemas existentes não mudam — o WeOS se adapta a eles.

json
# Exemplo: Conexão com Salesforce via API WeOS
POST /api/v2/public/connections
{
  "type": "salesforce",
  "name": "CRM Produção",
  "credentials": {
    "instance_url": "https://company.salesforce.com",
    "access_token": "<token>"
  }
}

Fluxos

O editor visual de fluxos permite que qualquer gestor desenhe como os agentes devem se comportar. Aprovações, condições, bifurcações e escalonamentos — tudo no diagrama, sem código. Versionamento e rollback instantâneo.

Marketplace

A maior biblioteca de IA do mundo. Mais de 4.300 itens incluindo modelos de IA, servidores MCP, AIWares, nós de automação, módulos IoT e widgets. Todos disponíveis para adicionar ao seu workspace com um clique.

Architecture

WeOS is architected in layers that separate responsibilities: infrastructure, orchestration, execution and observability.

1

Infrastructure Layer

Cloud, On-Premise or Hybrid. Your data never leaves your environment.

2

Orchestration Layer

WeOS Runtime manages agents, sessions, runs and persistent memory.

3

Execution Layer

Agents execute tasks using LLMs, tools and connections.

4

Observability Layer

Logs, metrics, costs and immutable audit of every action.

Deployment Modes

Managed Cloud

Infrastructure managed by WeGen-AI. Zero config, 99.9% SLA, automatic scaling.

On-Premise

Deploy in your datacenter or private cloud. Data never leaves your environment. Ideal for regulated industries.

Hybrid

Orchestration in WeGen cloud, agent execution in your environment. Maximum flexibility.

Best Practices

Recommendations for teams implementing WeOS in production.

Agent Naming

Use descriptive names indicating department, function and version. E.g. `sales-sdr-qualification-v2`. Makes observability and auditing easier.

Memory Management

Separate short-term (session) and long-term (workspace) memory. Avoid passing sensitive data in agent context when local storage is an option.

RBAC & Permissions

Apply the principle of least privilege. Each agent should only have access to the connections and knowledge bases needed for its function.

Audit Trail

Enable full audit on flows involving financial, health or legal data. The immutable trail is a compliance requirement for regulated industries.

Flow Testing

Use the staging environment to test flows with synthetic data before production. Version flows and keep rollback available.

Cost Control

Set per-agent cost alerts in the Observability module. Define token limits per session for high-volume agents.

API Reference v2

The WeOS v2 public API allows integrating the platform with any external system. All operations go through the API — including AIWare execution.

Base URL

url
https://api-weos.wegen-ai.com/api/v2/public

Authentication

All requests require Bearer token authentication. Get your token in the WeOS dashboard under Settings → API Keys.

http
Authorization: Bearer <WORKSPACE_PUBLIC_TOKEN>

Auth

Authentication and workspace switching.

Chat & Completions

OpenAI-compatible endpoints for completions and responses.

Estes endpoints são compatíveis com o formato OpenAI — você pode usar SDKs OpenAI apontando para a URL base do WeOS.

Agents

Management and execution of WeOS agents.

Sessions

Sessions maintain context for multi-turn conversations with agents.

Runs

Runs are individual executions of agents or flows. Each run has a traceable ID.

Models, Tools & MCP

Access to the catalog of models, tools and MCP servers available.

Multimedia

Image generation, speech synthesis and audio transcription.

Router & Workflows

Intelligent request routing and workflow execution.

Health

API status check.

AIWare Documentation

AIWares are configured and operated entirely through WeOS. There is no separate API for AIWares — all integration uses the `/agents`, `/sessions` and `/runs` endpoints from API v2.

How to integrate an AIWare via API

01

List workspace agents (`GET /api/v2/public/agents`) to get the AIWare agent's ID.

02

Create a session or use `run` directly if you don't need message history.

03

Execute the agent via `POST /api/v2/public/agents/{id}/run` or `/stream` for streaming.

bash
# 1. Listar agentes (encontrar ID do AIWare)
curl https://api-weos.wegen-ai.com/api/v2/public/agents -H "Authorization: Bearer <token>"

# 2. Executar o AIWare via agents/{id}/run
curl -X POST https://api-weos.wegen-ai.com/api/v2/public/agents/agent_sdr_abc/run \
  -H "Authorization: Bearer <token>" \
  -H "Content-Type: application/json" \
  -d '{"input": "Lead: Maria Silva, Diretora, 200 funcionários"}'

Configuration via WeOS

AIWares are configured in the WeOS dashboard without code. The visual editor allows:

  • Define the agent's persona and instructions
  • Connect knowledge bases and external connections
  • Configure decision and approval flows
  • Define access permissions by department
  • Enable auditing and alerts

Usage Policy

Use of the WeOS platform is subject to the following terms and guidelines.

Acceptable Use

  • Automation of internal business processes
  • Customer service with configured human oversight
  • Analysis of the organization's proprietary documents and data
  • Integration with third-party systems with proper authorization
  • Development and testing of AIWares in staging environment

Prohibited Use

  • Generation of misleading, defamatory or illegal content
  • Processing of minors' data without parental consent
  • Attempting to bypass platform governance controls
  • Using third-party credentials without authorization
  • Denial of service attacks or vulnerability exploitation

Data Retention

Execution logs are retained for 90 days on the Professional plan and a custom period on Enterprise. Session data can be configured for automatic retention.

Rate Limits

Limits vary by plan. The Starter plan allows 100 req/min. Professional and Enterprise have custom limits. HTTP 429 responses indicate rate limit reached — implement exponential backoff.

Privacy Policy

Last updated: January 2025

WeGen-AI Tecnologia LTDA is committed to protecting the personal data of its clients and users, in compliance with Brazil's General Data Protection Law (LGPD — Law No. 13,709/2018).

Data Collected

We collect only the data necessary to provide the service:

  • Registration data: name, email, company, role
  • Usage data: execution logs, performance metrics, per-agent costs
  • Infrastructure data: access IPs, authentication timestamps
  • Data processed by agents: as configured by the client (data in transit, not stored by default)

Purpose of Processing

  • Provision and continuous improvement of WeOS services
  • Technical support and service communication
  • Compliance with legal and regulatory obligations
  • Security and fraud prevention

Your Rights (LGPD)

  • Confirmation of existence of processing
  • Access to data
  • Correction of incomplete or incorrect data
  • Anonymization, blocking or deletion of unnecessary data
  • Data portability
  • Withdrawal of consent

Contact — DPO

To exercise your rights or clarify privacy questions, contact our Data Protection Officer:

privacy@wegen-ai.com