Frequently Asked Questions

Ottolax is an AI-powered office automation platform built by Titanium Labs LLC. It manages teams of AI agents that handle executive, operations, and assistant-level work — so your human team can focus on high-leverage activities.

The platform includes an agent management system, a Kanban-style task board, cron-based scheduled tasks, a Git integration, performance analytics, and a fully OpenAI-compatible API. It’s built on Claude Code CLI with an OpenAI-compatible FastAPI wrapper.

Each Ottolax agent is powered by Claude (Anthropic’s AI model) running through the Claude Code CLI. You define an agent’s role, skills, and description — and it uses that context to handle tasks appropriately.

Agents can work in two modes:

• Chat Mode — Conversational interface for everyday tasks
• Terminal Mode — Raw output for technical tasks like coding, scripting, and data processing

Agents receive tasks from the Kanban board, via the chat interface, or through scheduled triggers. They work asynchronously and report results back to the dashboard in real time via WebSocket.

Ottolax supports four model options:

Model	Description
sonnet	Claude Sonnet 4 — Best speed/intelligence balance (default)
opus	Claude Opus 4 — Most capable for complex agents and coding
haiku	Claude Haiku 4.5 — Fastest for simple, high-volume tasks
auto	Auto-select the best model based on the task

Ottolax also has an automatic fallback system: if your primary Claude CLI subscription runs out, it seamlessly switches to the Anthropic API using a backup API key — zero downtime for your agents.

Yes — Ottolax exposes a fully OpenAI-compatible API (/v1/chat/completions, /v1/responses, /v1/models). Any tool or SDK that works with OpenAI will work with Ottolax by simply changing the base_url to your Ottolax instance.

Compatible tools include LangChain, LobeChat, Continue.dev, Cursor, AnythingLLM, OpenWebUI, and your own Python or JavaScript applications using the OpenAI SDK.

Ottolax takes security seriously:

• JWT Authentication — Secure token-based auth with access tokens (15-min expiry) and refresh tokens (7-day expiry)
• Encrypted Storage — All sensitive credentials (API keys, OAuth tokens) are encrypted at rest using Fernet symmetric encryption
• Role-Based Access Control — Admin, User, and Viewer roles with granular permissions
• Audit Logging — Complete trail of all security-relevant operations
• Rate Limiting — 100 req/min per IP for API endpoints, 5 req/min for login
• Security Headers — HSTS, X-Frame-Options, X-Content-Type-Options, XSS protection

Credentials are never stored in plaintext and are encrypted using a key you control via the ENCRYPTION_KEY environment variable.

Yes — Ottolax is designed for self-hosting. Options include:

• Local — Run directly with python wrapper.py
• Docker — One-command Docker deployment
• Coolify — Deploy to your own Coolify instance
• Render — One-click Render deployment
• Any VPS — Works on any server running Python 3.11+

A Docker image is provided and the setup takes under 5 minutes. See our deployment documentation for step-by-step guides for each platform.

Early access members get:

• Full access to the Ottolax platform (agent management, task board, scheduled tasks, dashboard)
• Priority support from the Titanium Labs team
• Direct input into the product roadmap
• Locked-in pricing before public launch

To request early access, sign up on our contact page with your business email. We’ll reach out within 48 hours.

Ottolax has an automatic fallback system. It continuously monitors for subscription, quota, and rate limit errors from the Claude CLI. When detected, it seamlessly switches to the Anthropic API using your ANTHROPIC_API_KEY environment variable.

The fallback is:

1. Automatic — no manual intervention required
2. Logged — you can see when and how often fallbacks occur in your dashboard
3. Transparent — your agents continue working without interruption

You can also configure Ottolax to use the Anthropic API exclusively if you prefer direct API access over the CLI subscription model.

There is no hard-coded limit on the number of agents in the system. Practical limits depend on your infrastructure (server resources) and Claude API rate limits.

Most teams start with 3–10 agents — one orchestrator that delegates, and specialized workers for different domains (content, code, research, scheduling, etc.).

You can create, pause, and delete agents at any time from the dashboard. Agent conversation history is preserved even when an agent is idle.

No. The Ottolax dashboard is a fully visual interface — you can create agents, manage tasks, set schedules, and interact with your AI team without writing a single line of code.

If you want to integrate Ottolax into your own applications, a developer-friendly OpenAI-compatible API is available. But for day-to-day use, the web dashboard handles everything.

For self-hosted deployment, some basic command-line familiarity is helpful (setting environment variables, running Docker). Managed cloud hosting options that require zero setup are on the roadmap.