Your AI. Your hardware.

BlackRoad AI builds intelligence that runs on your devices. Lucidia is an AI assistant with persistent memory, local inference, and zero cloud dependency. Your conversations never leave your network.

Try Lucidia Explore products
52
TOPS local inference
16
AI models deployed
5
edge nodes
0
bytes sent to cloud

AI that lives where you do

Every BlackRoad AI product is designed to run locally first. Cloud is optional, never required. Your data stays on hardware you own, behind networks you control.

Lucidia Assistant

A conversational AI with persistent memory that remembers every interaction across sessions. Powered by Ollama with 16 models running on local hardware. Ask it anything — it knows your context because it lives in your workflow.

Sovereign Inference

52 TOPS of neural processing across Hailo-8 accelerators and Apple Silicon. Run CodeLlama 34B, Mixtral 8x7B, and DeepSeek locally with sub-second latency. No API keys, no rate limits, no per-token billing.

Memory System

SQLite-backed persistent memory with FTS5 full-text search, knowledge graph extraction, and cross-session context. Lucidia remembers solutions, patterns, and preferences. Every session builds on the last.

Multi-Device Sync

WireGuard-encrypted sync across all your devices. Start a conversation on your Mac, continue on your Pi cluster, pick it up on your phone. End-to-end encrypted, peer-to-peer, no relay servers.

Agent Fleet

Deploy specialized AI agents for code review, research, writing, and system administration. Each agent has its own memory, personality, and skill set. Coordinate through RoundTrip sovereign chat.

Voice Interface

Talk to Lucidia with local speech-to-text and text-to-speech. Whisper runs on-device for transcription. No audio ever leaves your hardware. Works offline, works air-gapped, works everywhere.

Pricing

Choose the tier that fits your workflow. All plans include local inference and persistent memory.

Rider

$29/mo
  • Lucidia AI assistant
  • Persistent memory (unlimited)
  • 3 AI models included
  • Single-device deployment
  • Community support
Get started

Enterprise

$299/mo
  • Everything in Paver
  • Unlimited agents
  • Team memory sharing
  • Custom model fine-tuning
  • Air-gapped deployment
  • Dedicated support + SLA
Contact us