Olympian Motors is building next-gen EVs that challenge our notions of what EVs look like! As part of their experience, they wanted to deliver a next-gen seamless, distraction-free driving experience. But removing traditional screens meant they had to build a natural, hands-free way for drivers to monitor and control their vehicles. Olympian Motors partnered with GoML to build a voice-powered AI solution that provides real-time vehicle information, executes commands, and responds to queries, entirely screen-free. GoML built an AI powered solution based on our proprietary accelerated AI delivery framework and LLM boilerplates.
The problem: lack of intuitive, hands-free vehicle control
Olympian Motors’ vehicles eliminated traditional dashboards, which made interacting with the vehicle a self-imposed challenge. Drivers could not easily access vital information such as battery levels, tire pressure, fluid levels, location, speed, or traffic conditions. Controlling vehicle functions like door locking/unlocking, temperature adjustment, or switching driving modes requires complex manual workflows or fragmented apps.
Existing systems lacked conversational context, contextual recommendations like charging range, and multi-turn interactions.
The absence of an intelligent, voice-first solution meant drivers faced reduced convenience, higher cognitive load, and a less safe driving experience. The need for an in-cabin AI assistant capable of understanding natural language, executing commands in real time, and providing context-aware responses was never more critical.
The solution: AI-powered in-cabin voice assistant
GoML built a cloud-based in-cabin voice assistant integrated with Olympian Motors’ 12 API endpoints. Using advanced large language models (LLMs) mediated via Amazon Bedrock, the assistant interprets natural language commands, maintains conversational context, and delivers intelligent, context-aware responses. It enables real-time monitoring of vehicle status, execution of control commands, and contextual recommendations, such as estimated range and charging guidance. A minimalistic web interface was developed for POC demonstration to validate the voice-to-voice interaction pipeline and API integration.
AI voice assistant for every driver-vehicle interaction
- Conversational agent: Processes natural language voice commands and manages context-aware dialogue.
- Query agent: Retrieves real-time vehicle data via Olympian Motors’ APIs for information like battery levels, tire pressure, and location.
- Control agent: Executes vehicle commands including door locking/unlocking, temperature adjustment, and driving mode selection.
- Context agent: Provides intelligent, contextual responses such as range estimates, charging recommendations, and multi-turn conversational continuity.
Powered by Amazon Bedrock + GoML AI Matic
- Language intelligence: Claude 3.5 via Amazon Bedrock interprets natural language and coordinates between agents.
- Voice interface stack: Speech-to-text and text-to-speech pipelines enable hands-free communication.
- API integration: 12 Olympian Motors endpoints provide real-time vehicle data and command execution.
- Multi-agent framework: Modular agents orchestrate conversational and command workflows for seamless vehicle interaction.
Scalable architecture across cloud-native components
- Data ingestion: AWS API Gateway and Python preprocessing pipelines for real-time vehicle data.
- Data enrichment: Contextual embeddings and prompt engineering for intelligent responses.
- Knowledge layer: Vector databases store session context and conversation history.
- LLM orchestration: RAG workflows via Lambda functions ensure dynamic command and query handling.
- CI/CD ready: Full deployment and update pipeline via AWS CodePipeline.
The impact: safer and smarter driving with AI
Olympian Motors’ drivers can now interact with their vehicles using natural language and receive real-time responses. Manual intervention is no longer needed for monitoring or controlling vehicle functions.
- Reduction in manual effort: Automated voice interactions replace fragmented control workflows.
- Simplified vehicle interaction: Context-aware AI reduces cognitive load and improves user experience.
Lessons for automotive innovators
What Olympian Motors learned:
- Screenless vehicles require voice-first, AI-powered control systems.
- Contextual understanding and conversational continuity enhance safety and usability.
- Integrating AI assistants with existing APIs demands robust orchestration and error handling.
- Continuous testing ensures accuracy and reliability of real-time vehicle interactions.
Advice for automotive teams:
- Identify repetitive queries and controls suitable for voice automation.
- Use modular AI agents to manage commands, queries, and contextual responses.
- Pair LLMs with vector databases for session context and multi-turn dialogue.
Looking to make your systems truly voice-enabled?
Let GoML help you build voice assistants for your environments so that your users can get a natural and hands-free interaction experience.
