OEM Integration FAQ

Common questions about integrating Soro's plug-in voice brain modules into appliances, toys, and medical devices.

General Questions

Soro is a plug-in hardware module that adds natural speech understanding to any appliance, toy, or medical device. Our pre-certified module runs entirely offline with wake-word detection, Whisper-grade speech-to-text, and a 350M-parameter LLM that outputs JSON intent. Perfect for OEMs who need voice capabilities but lack AI talent.

New privacy laws like GDPR, HIPAA, and COPPA block cloud microphones in toys, medical devices, and appliances. Soro's offline processing eliminates data collection concerns while delivering <300ms response times. Your customers' voices never leave the device, ensuring complete privacy compliance by design.

OEMs go from idea to working prototype in a weekend. Our drop-in modules have standard UART/SPI/I2C interfaces and simple JSON output format. Complete integration documentation, sample code, and engineering support help you skip months of AI development. Most customers have voice-enabled prototypes running within 48 hours.

Soro modules are pre-certified for FCC Part 15, CE marking, UL safety, and RoHS compliance. This eliminates months of regulatory approval for your products. We also provide FDA pathway guidance for medical devices and maintain compliance documentation for GDPR, HIPAA, and COPPA requirements.

Technical Questions

Our modules feature ARM Cortex-A78 processors, 8GB LPDDR5 memory, and integrated neural processing units. Available in standard form factors for easy integration. Ultra-low power design with wake-word detection consuming <1mW. Standard interfaces include UART, SPI, I2C, plus optional WiFi/Bluetooth connectivity.

Soro modules communicate via UART, SPI, or I2C with simple JSON command/response format. No complex APIs or cloud dependencies. Standard GPIO pins for device control, analog audio I/O, and power management. We provide reference designs, sample code, and integration libraries for popular microcontrollers and development boards.

Absolutely! We can train custom wake words like "Hey [YourBrand]" or product-specific activation phrases. Custom vocabulary and domain-specific commands are also available. Training typically takes 2-3 weeks and includes optimization for your specific acoustic environment. Each module can store multiple wake word models for different products.

Wake-word detection runs continuously on <1mW power consumption. Full voice processing during active conversations uses 2-5W depending on complexity. Sleep/standby modes available for battery-powered devices. Optimized power management extends battery life while maintaining always-listening capabilities. Perfect for portable and battery-operated products.

Security & Privacy

Security is our top priority. All audio data is encrypted with AES-256 during transit (TLS 1.3) and at rest. We implement zero data retention by default - audio is processed in real-time and immediately deleted. Our infrastructure is SOC 2 Type II certified, ISO 27001 compliant, and undergoes regular security audits. We also support private cloud deployments for maximum security control.

Yes, Soro is fully GDPR compliant with data processing agreements (DPA) available for EU customers. We're also HIPAA compliant for healthcare applications with Business Associate Agreements (BAA) available. Our platform supports data residency requirements, right to deletion, data portability, and all other privacy regulations. We maintain detailed audit logs for compliance reporting.

By default, Soro does not store or retain any audio data. Audio is processed in real-time and immediately deleted after analysis. Only the extracted insights (text transcripts, classifications, etc.) are temporarily cached for performance optimization and can be configured for immediate deletion. Customers can optionally enable secure storage for features like model training, but this requires explicit consent and configuration.

Integration & Support

Most developers can integrate Soro's basic functionality in under 30 minutes using our SDKs and comprehensive documentation. Simple API calls can be implemented in just a few lines of code. For more complex integrations involving real-time streaming or custom models, typical implementation time is 1-2 days. Our solutions engineers provide free integration support to help you get started quickly.

We offer multiple support tiers: free community support via our developer forum, email support for paid customers, and dedicated technical account managers for enterprise clients. Our documentation includes API references, tutorials, sample code, and troubleshooting guides. Enterprise customers receive priority support with guaranteed response times, phone support, and direct access to our engineering team.

Absolutely! Soro is designed for flexible integration with existing systems. We provide REST APIs, webhooks, and streaming protocols that work with any technology stack. We also offer pre-built integrations for popular platforms like Salesforce, Slack, Microsoft Teams, and major cloud providers (AWS, Azure, GCP). Custom integrations and on-premise deployments are available for enterprise customers.

Yes, we offer comprehensive professional services including custom integration development, model training and optimization, system architecture consulting, and ongoing managed services. Our team of audio AI experts and solutions engineers can help with everything from initial proof-of-concept to full-scale production deployment. Professional services are typically included in enterprise packages or available as add-on services.

Still Have Questions?

Can't find the answer you're looking for? Our team is here to help with any questions about Soro's audio intelligence platform.

Contact Support Schedule a Demo