Cars are lasting longer
The average vehicle on U.S. roads is 13 years old. Modern vehicles routinely last hundreds of thousands of miles, and owners are choosing to upgrade the cars they already love instead of replacing them.
An AI voice agent platform for the car.
Open Agent (Invite Only)
Live OBD II Onboard Data
Reads OBD data from any car and turns it into conversation.
Custom Personalities
Customizable AI car personas, matched to the car by VIN.
Tools platform
Uses device capabilities and both public and OAuth APIs.
Demo Drive
Test users already lining up
The Prototype Everyone Wants
Ready to go from 1 to 20 or more test units immediately
A test fleet would go live almost overnight: everyone who knows about Talking Car Steve has asked for their own.
- A Bluetooth OBD adapter (OBDLink MX+)
- A dedicated, data-connected Android device
Why Now?
Voice: the UX of the future
AI is changing how we interact with tech — not through apps and menus anymore.
Cars already have computers
Since 1996, virtually every passenger vehicle has standardized onboard telemetry through the OBD-II interface. Every modern car already knows its speed, RPM, and hundreds of other data points - it simply has no intelligent way to communicate them.
The opportunity: Upgrade the experience—not the car
Talking Car Steve gives existing vehicles an intelligent AI upgrade. A lightweight hardware device paired with a software subscription transforms today's cars into context-aware conversational agents that understand the vehicle, the journey, and the people inside.
Built for the Next Generation of Vehicles
The same platform naturally extends beyond today's drivers. Fleet operators gain intelligent vehicle assistants. Automakers gain configurable AI personalities and branded experiences. And as autonomous vehicles become commonplace, conversation will become the primary interface between passengers and the vehicle itself.
The Product Vision
From personal cars to autonomous fleets
The product begins as a consumer and enterprise sales driving companion, then expands through fleet deployments, OEM integrations, and conversational experiences for autonomous vehicles. Each version uses the same web UI with underlying platform configured for specific vehicles and/or environments and reciprocol set of tools.
Consumers
Enterprise Outside Reps
Fleet Vehicles
OEM Automakers
Autonomous Cars
Architecture
More Than an App
A Powerful, Hands-Free AI Agent
Talking Car Steve brings realtime car data together with voice AI, tools, realtime awareness and memory - turning a physical machine into an intelligent agent.
Knows the whole context
Steve understands the vehicle, the route, the passengers, the destination, and what is happening around you. It can help with driving, productivity, planning, entertainment, and vehicle questions without taking your eyes off the road.
Conversation becomes action
Ask for a destination, compare nearby places, change plans, or continue a trip you started on the web. Steve can turn the conversation directly into hands-free navigation while you keep your attention on the road.
Get things done and enjoy the ride
Let Steve handle things like finding contacts, logging sales calls, pulling news and information, or simply turning on your favorite musical playlist.
Choose who you drive with
The agent can be Steve, or Elizabeth - or any persona you want: a calm copilot, a knowledgeable mechanic, an energetic road-trip companion, a storyteller. For every user VIN you get to decide what character comes alive with AI.
The agent travels with you
Steve can begin helping on the web, continue on a mobile device, and become fully hands-free inside the car. Your plans, memory, tools, and conversation move across environments instead of restarting each time.
A platform that keeps expanding
Navigation, diagnostics, weather, messaging, calendars, research, entertainment, local search, and third-party services are exposed through a configurable tool system. New capabilities can be added without rebuilding the core conversational experience.
Voice ID and Profiles
The agent doesn't just hear you — it knows you. Voice Activity Detection catches the moment speech begins, and on-device speaker recognition figures out who said it, all before the first word reaches the AI. Enroll multiple passengers and it greets each of you by name, adapts its tone, and keeps a shared conversation going across the whole car.
Voice Activity Detection
A device-local neural VAD catches the instant someone starts speaking — no push-to-talk, no wake word. It discards road noise and music bleed so the AI only ever hears real speech.
Voice Enrollment
Each passenger records short voice samples. That print lives on-device and lets Steve match voices in real time from that moment on.
Multi-Person Recognition
Passenger says something? Steve answers them directly. Driver asks a follow-up? He knows it's you. The whole car can have a conversation, not just the person who opened the app.
What Steve Can Do
The tools Steve can reach for mid-conversation, grouped by what they touch.