Frequently Asked Questions
What is Hoziron?
Hoziron is an enterprise AI agent infrastructure platform. It provides a model-agnostic runtime for deploying, managing, and orchestrating AI agents with built-in data sovereignty, compliance guardrails, and multi-provider support. It's a compliance-ready distribution targeting the insurance vertical.
What LLM providers are supported?
Out of the box: Anthropic (Claude), OpenAI (GPT), Groq (Llama, Mixtral), and any Ollama or vLLM instance. Since Ollama and vLLM expose OpenAI-compatible APIs, any model they serve is available to Hoziron.
Can I use local models?
Yes. Configure Ollama or vLLM as a provider and disable cloud providers. This is the recommended setup for air-gapped deployments. See the local models guide.
How do I add a new provider?
Add a [providers.<name>] section to your config.toml with the API key env var and base URL. If the provider exposes an OpenAI-compatible API (like vLLM, LiteLLM, or Azure OpenAI), configure it as the openai provider with a custom base_url.
What's the difference between a skill and a competency?
A skill is a tool the agent can invoke (query a database, call an API, read a file). A competency is a complete behavior definition — it bundles a system prompt, required skills, configuration, and metrics. Think of skills as individual capabilities and competencies as complete job descriptions.
How does PII protection work?
Data flowing between agents in multi-agent workflows passes through a PII tokenization pipeline. Detected PII (names, SSNs, policy numbers) is replaced with opaque tokens before crossing agent boundaries. Each agent has its own isolated memory scope — no cross-agent memory access is possible.
Can agents communicate with each other?
Not directly. Agents are isolated by design. They communicate through workflow steps — the workflow engine passes outputs from one agent as inputs to another, with PII tokenization applied at each boundary. This ensures data sovereignty and prevents information leakage.
Is my data sent to cloud LLM providers?
Only the messages you send to agents. Hoziron routes agent conversations to whichever LLM provider you configure. If you use local models (Ollama/vLLM), no data leaves your network. The platform itself stores all state locally in SQLite — nothing phones home.
How do I run Hoziron air-gapped?
- Use local models (Ollama or vLLM) instead of cloud providers
- Disable cloud providers in config (
enabled = false) - Transfer packages offline using
hoziron catalog export/hoziron catalog import
See the air-gapped deployment guide.
What's the performance overhead?
Minimal. Each agent uses ~10 MB base memory plus conversation context. CPU is negligible when idle and bursts during LLM interactions. The platform is a single Rust binary — no JVM warmup, no garbage collection pauses. A 2-core, 2 GB machine can run 5+ agents comfortably.
How many agents can I run?
Depends on your hardware and conversation volume:
| Resources | Agents |
|---|---|
| 2 cores, 2 GB | 1–5 |
| 4 cores, 4 GB | 10–50 |
| 8 cores, 8 GB | 50–200 |
| 16+ cores, 16+ GB | 200+ |
Idle agents have negligible overhead. Active agents consume resources proportional to conversation length and LLM call frequency.
How do I back up my data?
All state lives in $HOZIRON_HOME/data/. For a clean backup:
hoziron stop
cp -r $HOZIRON_HOME/data/ /backup/hoziron-$(date +%Y%m%d)/
hoziron start
For live backups (SQLite supports concurrent reads):
sqlite3 $HOZIRON_HOME/data/hoziron.db ".backup '/backup/hoziron.db'"
Can I migrate from another platform?
There's no automated migration tool yet. However, since Hoziron agents are defined by TOML manifests and competency definitions, you can recreate your agent configurations manually. Data integrations use standard protocols (PostgreSQL, REST APIs, S3), so reconnecting data sources is straightforward.
What happens when I delete an agent?
Deletion removes the agent from the registry, clears its conversation history, and frees its memory scope. The agent must be in Terminated state before deletion (stop it first). Audit trail entries for the agent's activity are preserved.
How do I contribute a package?
- Create your package:
hoziron package init --type <type> --name <name> - Build:
hoziron package build ./my-package/ - Lint:
hoziron package lint ./my-package/ - Sign:
hoziron package sign ./my-package/ - Publish:
hoziron package publish ./my-package/
See the publishing guide for full details.
Related: