diff --git a/DEPLOY_SERVIDOR.md b/DEPLOY_SERVIDOR.md
new file mode 100644
index 0000000..0156df0
--- /dev/null
+++ b/DEPLOY_SERVIDOR.md
@@ -0,0 +1,121 @@
+# Deploy no Servidor da Empresa (Debian/Ubuntu + systemd)
+
+## 1) Fluxo correto com Git
+
+Sim: primeiro voce faz commit e push da sua maquina local, depois no servidor faz pull.
+
+Exemplo:
+
+```bash
+# local
+git add .
+git commit -m "ajusta mysql/tools + telegram webhook"
+git push origin main
+
+# servidor
+cd /opt/orquestrador
+git pull origin main
+```
+
+## 2) Preparacao no servidor
+
+```bash
+cd /opt/orquestrador
+python3 -m venv venv
+source venv/bin/activate
+pip install -U pip
+pip install -r requirements.txt
+```
+
+## 3) Configurar .env.prod
+
+Crie/atualize o arquivo `.env.prod` no servidor:
+
+```env
+GOOGLE_PROJECT_ID=seu-projeto
+GOOGLE_LOCATION=us-central1
+VERTEX_MODEL_NAME=gemini-2.5-flash
+
+DB_HOST=127.0.0.1
+DB_PORT=3306
+DB_USER=root
+DB_PASSWORD=sua_senha
+DB_NAME=orquestrador_mock
+
+MOCK_DB_HOST=127.0.0.1
+MOCK_DB_PORT=3306
+MOCK_DB_USER=root
+MOCK_DB_PASSWORD=sua_senha
+MOCK_DB_NAME=orquestrador_mock
+
+AUTO_SEED_TOOLS=true
+AUTO_SEED_MOCK=true
+MOCK_SEED_ENABLED=true
+
+# telegram (opcional)
+# TELEGRAM_BOT_TOKEN=...
+# TELEGRAM_WEBHOOK_SECRET=...
+```
+
+## 4) Credencial para Vertex AI
+
+Defina um dos formatos abaixo no servidor:
+
+- `GOOGLE_APPLICATION_CREDENTIALS=/opt/orquestrador/sa.json`
+- ou Application Default Credentials (`gcloud auth application-default login`)
+
+A service account deve ter permissao no Vertex AI (ex.: `roles/aiplatform.user`).
+
+## 5) Configurar service do systemd
+
+1. Copie o template e ajuste usuario/path:
+
+```bash
+sudo cp deploy/systemd/orquestrador.service.example /etc/systemd/system/orquestrador.service
+sudo nano /etc/systemd/system/orquestrador.service
+```
+
+2. Recarregue e inicie:
+
+```bash
+sudo systemctl daemon-reload
+sudo systemctl enable --now orquestrador
+sudo systemctl status orquestrador
+```
+
+3. Logs:
+
+```bash
+journalctl -u orquestrador -f
+```
+
+## 6) Testes rapidos
+
+```bash
+curl -s http://127.0.0.1:8080/openapi.json | head
+```
+
+```bash
+curl -s -X POST http://127.0.0.1:8080/chat \
+  -H "Content-Type: application/json" \
+  -d '{"message":"Quero um sedan ate 50000"}'
+```
+
+```bash
+curl -s -X POST http://127.0.0.1:8080/mock/consultar-estoque \
+  -H "Content-Type: application/json" \
+  -d '{"preco_max":50000,"categoria":"sedan"}'
+```
+
+## 7) Atualizacao em producao
+
+Sempre que subir novas mudancas:
+
+```bash
+cd /opt/orquestrador
+git pull origin main
+source venv/bin/activate
+pip install -r requirements.txt
+sudo systemctl restart orquestrador
+sudo systemctl status orquestrador
+```
diff --git a/deploy/systemd/orquestrador.service.example b/deploy/systemd/orquestrador.service.example
new file mode 100644
index 0000000..1551728
--- /dev/null
+++ b/deploy/systemd/orquestrador.service.example
@@ -0,0 +1,21 @@
+[Unit]
+Description=AI Orquestrador API (FastAPI/Uvicorn)
+After=network.target
+
+[Service]
+Type=simple
+User=vitor
+Group=vitor
+WorkingDirectory=/opt/orquestrador
+EnvironmentFile=/opt/orquestrador/.env.prod
+Environment=PATH=/opt/orquestrador/venv/bin
+ExecStart=/opt/orquestrador/venv/bin/uvicorn app.main:app --host 0.0.0.0 --port 8080
+Restart=always
+RestartSec=5
+
+# hardening basico
+NoNewPrivileges=true
+PrivateTmp=true
+
+[Install]
+WantedBy=multi-user.target