Quick Start: Connecting and Invoking Custom AI Models

1. Overview & Technical Architecture#

Connect and test the integrated custom model in your SmartERP deployment.

Prerequisites#

• SmartERP instance running on Python 3.12+ and PostgreSQL 16.
• Model service running locally or on network server (e.g., http://localhost:11434 or https://ai.smarterp.et/v1).

1. Verifying Local Model Health#

Open a terminal and verify the model endpoint using curl:

curl -s http://localhost:11434/api/tags | jq .

Verify that smarterp-sovereign:latest appears in the active models list.

{screenshot:custom-ai-quickstart-terminal}

2. Configuring the SmartERP Model Connector#

1. Log into SmartERP as an Administrator.
2. Go to Settings → Technical → Ashewa AI → Model Endpoints.
3. Click New and fill in:
   • Endpoint Name: Local Sovereign AI Server
   • Provider Protocol: Ollama / OpenAI-Compatible REST
   • Base URL: http://localhost:11434/v1
   • Model Identifier: smarterp-sovereign:latest
4. Click Test Endpoint Latency. Confirm HTTP 200 with response latency under 350ms.

{screenshot:custom-ai-quickstart-odoo-config}

3. Testing Generation#

Open the floating AI composer, select Local Sovereign AI Server, and submit a test prompt:

"What are the mandatory tax deductions under Ethiopian Proclamation 979/2016?"

Verify structured, compliant response.

2. Prerequisites & Infrastructure Requirements#

• Runtime Environment: Python 3.12+, PostgreSQL 16 with pgvector extension.
• Security Access Group: Ashewa AI / Administrator (Technical ID: muk_ai.group_ai_manager)
• Hardware Profile: Minimum 16 GB System RAM (CPU inference) or NVIDIA GPU with 8+ GB VRAM.

3. Step-by-Step Configuration & Verification#

Step 1: Navigating to Model Management#

1. Access your SmartERP instance at https://smarterp.et/odoo.
2. Open Settings → Technical → Ashewa AI → Model Registry.
3. Select the custom model profile.

{screenshot:30-quick-start-step1-navigation}

Step 2: Testing Endpoint Health & Response Verification#

1. Click the Run Diagnostic Health Check button.
2. Verify token generation speed (tokens/sec) and latency benchmarks.
3. Review sample response in the diagnostic console.

{screenshot:30-quick-start-step2-verification}

4. Advanced Configuration Points#

• Automated Tasks: muk_ai.ir_cron_model_health_monitor runs every 15 minutes.
• Multi-Tenant Boundaries: Custom model fine-tuning adapters are strictly segregated per tenant database.

5. Comprehensive Troubleshooting Matrix#