Most AI tools today are purely reactive: they sit idle until you type a prompt, answer in isolation, and forget everything once you close the tab. On top of that, proprietary platforms store your private conversations and projects in their own centralized servers.
LIGHT LPCA is built differently:
- Proactive, Not Just Reactive: Equipped with a background task engine (Shadow Brain) that can evaluate context, follow up on tasks, and prepare insights autonomously.
- Sovereign Privacy (BYODB): Your chat logs, memory facts, and personal history are stored in your own database (Neon PostgreSQL or MongoDB). No proprietary lock-in.
- Zero Markup (BYOK): Connect your own API keys (Groq, Google Gemini, OpenAI, Claude, OpenRouter, DeepSeek, or local Ollama). Use free tier keys from Groq or Gemini without paying subscription markups.
- Live Web & Reader Tools: Built-in web search and deep page scraping directly integrated into the reasoning loop.
A background cognitive engine that can wake up independently, organize your active tasks, summarize long-term context, and proactively help you continue where you left off.
- You maintain total control: the Shadow Brain can be toggled on or off at any time with a single click.
Connect the model providers you already use. Switch models on the fly depending on the task:
- High-Speed & Free Inference: Groq (
llama-3.3-70b-versatile) and Google Gemini (gemini-1.5-flash,gemini-2.0-flash). - Frontier Reasoning: OpenAI (GPT-4o, o-series), Anthropic Claude, DeepSeek, and OpenRouter.
- Local Models: Connect local Ollama or custom OpenAI-compatible endpoints directly.
True ownership of your AI context.
- Connect your own cloud database URI — whether PostgreSQL (like Neon or Supabase) or MongoDB Atlas.
- LIGHT automatically provisions and migrates required schemas in isolated, encrypted namespaces.
- Your chat records, persona memory, and embeddings never live in a shared central repository.
No waitlists, no invite codes, and no payment required to get started.
- Open the Platform: Visit lightlpca.com.
- Create Your Account: Sign up instantly.
- Connect Your Provider:
- Head over to Settings → API Keys.
- Paste a free key from Google AI Studio or Groq Console.
- Connect Your Database (Optional but Recommended):
- Provide a free Neon PostgreSQL or MongoDB Atlas connection string for full data sovereignty.
- Start Chatting: Ask questions, search the live web, or let the Shadow Brain assist your workflow.
| Provider | Recommended Models | Key Cost |
|---|---|---|
| Groq | llama-3.3-70b-versatile, llama-3.1-8b-instant |
Free Tier Available |
| Google Gemini | gemini-1.5-flash, gemini-2.0-flash, gemini-1.5-pro |
Free Tier Available (1500 req/day) |
| OpenAI | gpt-4o, gpt-4o-mini |
Pay-as-you-go |
| Anthropic | claude-3-5-sonnet-latest, claude-3-5-haiku-latest |
Pay-as-you-go |
| DeepSeek | deepseek-chat |
Very Low Cost |
| OpenRouter | Any supported aggregator model | Variable |
| Local / Custom | Ollama, vLLM, LM Studio | Free (Self-hosted) |
LIGHT is actively evolving. If you encounter an issue or have an architectural suggestion:
- Open an issue right here on GitHub.
- Share thoughts or suggestions to help shape the future of autonomous personal agents.