A trustworthy, verifiable personal AI assistant that brings institutional-grade analytical power to individuals, powered by a knowledge graph you own.
Not another AI chatbot. A sovereign knowledge architecture where your intelligence grows with you.
Your data is encrypted with keys only you hold. The server stores ciphertext. We cannot read your data even if we wanted to. This is architecture, not policy.
KalaBodha builds a personal knowledge graph of your life — contacts, relationships, decisions, patterns. It grows with every interaction. Facts live in the graph, not in model weights.
Run with Ollama locally for complete privacy. Connect to Claude, GPT, or any API for maximum capability. Swap models without losing a byte of data. The graph is the brain — the model is the voice.
Runs on your hardware. Docker Compose. No cloud required. Works offline with a local model. Your assistant, your machine, your rules.
Every action is logged in an append-only, cryptographically hashed audit trail. When connected to Guardian Mesh, actions pass through formally verified deterministic enforcement.
Fork it. Audit it. Extend it. The code IS the methodology. No hidden algorithms. No opaque model weights making decisions about your life.
KalaBodha brings a new approach to personal AI - trustworthy, verifiable, and built around the knowledge you already have.
| Traditional Approach | KalaBodha | |
|---|---|---|
| Data sovereignty | Cloud-hosted data | Your data stays on your hardware, encrypted with your keys. |
| Intelligence | Model-centric intelligence | Knowledge-graph-centric - structured, auditable, yours |
| Learning | Session-based context | Graph grows with every interaction |
| Model | Fixed model choice | Any model. Local or cloud. Swappable. |
| Verifiable answers | Model-generated answers | Answers are traceable with provenance chain. VERIFIED or OUT OF SCOPE. |
| Offline | Cloud required | Full offline capability with local LLM |
| Your data trains their model? | Varies by provider | Your data never enters any training pipeline - by architecture. |
| Open source | Proprietary | AGPL-3.0. Fully auditable. The code IS the methodology. |
| Cost | Subscription model | Free. Your hardware, your compute. Paid backup optional. |
Knowledge graph as primary intelligence. LLM as communication helper. Trustworthy, verifiable, yours.
We're building trustworthy, verifiable AI beyond models and agents. Looking for builders who want to create something meaningful.
Agentic loop, encryption layer, audit trail, CLI, Tauri desktop shell, module bridges
Email/calendar integration, OAuth, Neo4j graph operations, FastAPI, SHACL validation
Chat interface, graph browser, calendar view, approval workflow, Capacitor mobile
OWL 2 ontology design, graph population pipeline, Entity-State versioning, SPARQL
LoRA/QLoRA fine-tuning, DPO alignment, GGUF quantization, evaluation benchmarks
Prompt injection defense, graph poisoning prevention, formal verification (Lean 4)
Architecture docs, setup guides, API docs, tutorials, onboarding content
Interface design for everyday people, onboarding flow, graph visualization, brand
Non-negotiable. Structural, not policy.
No interview. No coding test. Show us your work, tell us what excites you, and let's build.
Rust, Python, TypeScript, Knowledge Graph, AI/ML, Security, Docs, or UX.
Title: "Collaborator: [Name] - [Area]". Short intro + link to your work.
We respond within 48 hours with a good first issue.