Toweringmedia AI Core for Magento 2
- Community
The Problem
- ✖ AI integrations are fragmented across multiple Magento extensions
- ✖ Provider credentials and model settings must be maintained separately
- ✖ Unexpected AI costs can occur without usage controls
- ✖ Scaling AI capabilities across the store becomes difficult and time-consuming
The Solution
Toweringmedia AI Core acts as the foundation for all Toweringmedia AI extensions, delivering centralized provider management, secure credential storage, shared AI services, and built-in cost controls for a scalable Magento AI ecosystem.
How It works
See It In Action
Toweringmedia AI Core for Magento 2
The shared AI infrastructure layer for all Toweringmedia extensions. Configure OpenAI, Google Gemini, Whisper, TTS, budgets, and safety gates once — Nora AI, Search Intelligence, and Product Builder inherit everything automatically.
One License. All AI Extensions.
Configure Once, Power Every AI Module
Toweringmedia AI Core is the shared mechanics layer that all AI-capable Toweringmedia extensions depend on. Enter your API key, pick a model, and set a daily budget once — Nora AI, Search Intelligence LLM, Product Builder, and future modules plug in automatically without re-implementing HTTP clients or safety gates.
- ✓ One license key activates the entire AI platform
- ✓ Master enable/disable for live API calls — safe for staging by default
- ✓ Dry Run mode: zero API spend, full local operation for testing
- ✓ License validated remotely — live calls blocked when subscription lapses
OpenAI-Compatible Provider
One API Key for All AI Features
Set your OpenAI-compatible API base URL, model, and key once. AI Core stores the key encrypted in core_config_data and routes all chat completions, summarization, and embeddings through this single configuration — no duplicated credentials across extensions.
- ✓ Works with OpenAI, Azure OpenAI, Ollama, or any compatible endpoint
- ✓ Encrypted key storage — never exposed in logs or admin exports
- ✓ Per-store model selection via store view scope
- ✓ Embeddings model field for semantic routing and exemplar seeding
Task Routing
Right Provider for Every AI Task
Assign different providers to different AI task types from one screen. Chat completions, summarization, image generation, speech-to-text, and text-to-speech can each use the best-fit provider without touching individual extension configs.
- ✓ Chat Completion — powers Nora AI conversations and agent tool calls
- ✓ Summarization — context compression and summary-tier memory
- ✓ Image Generation — Google Gemini for subject-consistent product photo editing
- ✓ Speech-to-Text — OpenAI Whisper for Nora voice input
- ✓ Text-to-Speech — OpenAI TTS for Nora voice responses
Voice — STT & TTS
Shared Voice Infrastructure for Nora AI
When Nora AI needs voice input or spoken responses, AI Core provides the shared Whisper STT and OpenAI TTS clients. No separate API key or config — both inherit from the Chat Provider settings above.
- ✓ OpenAI Whisper transcription via the same shared API key
- ✓ OpenAI TTS with model (
tts-1-hdortts-1) and voice selection - ✓ Master enable/disable for STT and TTS separately
- ✓ Leave both off until voice features are needed — zero cost impact
Production Safety
Built-In Spending Controls and Safety Gates
AI Core enforces call budgets, per-run caps, token budgets, and post-error cooldowns at the platform level — no external monitoring required. Google Gemini image generation is included for Product Builder's subject-consistent photo editing workflows.
- ✓ Daily Call Budget — max provider calls per store per UTC day (0 = block all)
- ✓ Max Calls Per Run — per-orchestration cap prevents runaway admin jobs
- ✓ Packet Token Budget — deterministic shrinking before any provider call
- ✓ Cooldown After Error — blocks calls for N seconds after a recorded failure
- ✓ Google Gemini — separate encrypted key for image generation tasks
Extensions That Depend on AI Core
AI Core is the required platform subscription for all Toweringmedia AI extensions.
Nora AI
Storefront AI customer service agent. Nora's multi-turn conversation, order lookup, voice input, and operator escalation all run through AI Core's agent runtime and memory.
Search Intelligence LLM
AI-powered SEO workflow orchestration. Uses AI Core's provider, packet budgeting, and call guards for safe, repeatable content generation at scale.
Product Builder
AI-assisted product configuration with Gemini image editing. Uses AI Core for chat completions and subject-consistent product photo generation via Google Gemini.
Questions about AI Core or dependent extensions?
Tell us which AI extensions you need, your Magento version, and any existing OpenAI or API setup. We reply within one business day.
Works Great With
Frequently Asked Questions
Does Toweringmedia AI Core work with any OpenAI-compatible provider?
Do I need AI Core to use Nora AI or Search Intelligence?
Does AI Core support daily call budgets and cost limits?
Is Toweringmedia AI Core compatible with Hyvä themes?
Can I test AI Core without spending on API calls?
📘 Documentation
Get started with Toweringmedia AI Core using our comprehensive documentation. Learn how to configure AI providers, manage models, set usage limits, and enable shared AI services across your Magento store.
- Step-by-step installation and setup instructions
- OpenAI, Gemini, Azure OpenAI, and Ollama configuration
- Voice services, embeddings, and model management
- AI budget controls and safety configuration
- Multi-store and extension integration guidance
💬 Support
Need assistance? Our Magento and AI specialists are available to help with installation, provider configuration, troubleshooting, and extension integration.
- Fast and reliable technical support
- Help configuring AI providers and models
- Troubleshooting API and connectivity issues
- Guidance on budgets, rate limits, and AI safety controls
- Regular updates and compatibility assistance
Toweringmedia AI Core is the foundational AI infrastructure module for Magento 2. It provides centralized configuration, provider management, security controls, and shared AI services that can be used across all Toweringmedia AI-powered extensions.
AI Core supports OpenAI, Azure OpenAI, Google Gemini, Ollama, and other OpenAI-compatible providers. You can configure provider credentials, models, and endpoints directly from the Magento admin panel.
Yes. AI Core includes budget controls, call limits, token restrictions, cooldown settings, and emergency disable options to help prevent unexpected AI costs and maintain operational control.
Yes. The extension supports speech-to-text and text-to-speech services, including OpenAI Whisper and OpenAI TTS, allowing connected extensions to deliver voice-enabled customer experiences.
AI Core is designed to power Toweringmedia AI extensions such as Nora AI, Search Intelligence, Product Builder, and future AI-enabled modules. These extensions can share the same provider configurations, services, and usage controls.
- Magento Version: Compatible with Magento 2.4.x (Open Source & Adobe Commerce)
- PHP Version: PHP 8.1, 8.2, and 8.3 compatible
- Framework: Built using Magento 2 native architecture (no core overrides)
- Installation Type: Composer installation supported
- Module Dependency: Required by Toweringmedia AI-powered extensions
- Supported AI Providers: OpenAI, Azure OpenAI, Google Gemini, Ollama, and OpenAI-compatible endpoints
- Embeddings Support: Yes
- Speech-to-Text: OpenAI Whisper support
- Text-to-Speech: OpenAI TTS support
- AI Task Routing: Provider-specific model assignment for different workloads
- Multi-Store Support: Yes
- Store Scope Configuration: Website / Store View level
- Credential Storage: Encrypted API key management
- Dry Run Mode: Available for testing and staging environments
- License Management: Centralized licensing for AI extensions
- Budget Controls: Daily limits, call restrictions, and token usage controls
- Security: Encrypted credentials and provider-level access controls
- Cron Support: Fully compatible with Magento cron jobs
- Admin Configuration Path: Stores → Configuration → Toweringmedia → AI Core
- Performance Impact: Minimal overhead with centralized AI service management