Solutions
Unified Speech-to-Text Gateway
Core components：GCP Apigee
Underlying support：GCP Cloud Functions (for logic distribution)
Solution Description：To address the varying needs of businesses regarding speech recognition accuracy and cost in different scenarios, and considering the limited service quotas on different cloud platforms, we have built a unified entry point based on the GCP API Gateway.
- Multi-cloud aggregation：Access Google Cloud Speech-to-Text, AWS Transcribe, and Azure Speech Service simultaneously through a single API interface.
- Automatic intelligent routing: Automatically route requests to the best-performing underlying platform based on language, audio quality, or response speed.
- Unified authentication and billing: Developers do not need to manage multiple platform keys; unified security authentication and traffic monitoring are handled through the GCP gateway.
Enterprise-grade large language model aggregation gateway
Core components：Global HTTP(S) Load Balancer，Managed Instance Groups，Instance Template，Cloud Memorystore，Cloud SQL
Underlying support：New API
Solution Description：Solution Description: To address the complexity of integrating multiple AI providers and the challenges of managing varying API protocols, we have developed a unified aggregation layer. Built on an elastic server architecture, this solution provides a standardized interface for top-tier LLMs including Gemini, GPT, and Claude,…LLMs, ensuring enterprise-grade scalability and cost-efficient AI operations.
- Unified API Protocol: Standardizes disparate LLM interfaces into a single OpenAI-compatible API, allowing developers to switch between models (e.g., Gemini to Claude) without modifying application code.
- Elastic Scaling & High Availability: Leverages GCP Managed Instance Groups (MIG) to automatically scale compute resources based on real-time traffic, ensuring low-latency responses even during peak demand.
- Intelligent Quota & Cost Management: Utilizes Cloud SQL and Redis to track token usage, manage user-level quotas, and implement response caching, significantly reducing operational costs and preventing budget overruns.
- Enterprise Governance & Security: Centralizes API key management via Secret Manager and provides comprehensive audit logs, enabling fine-grained access control and compliance monitoring across all AI interactions.
