A Container Native AI Platform by NetworkEvolution
Deploy AppTelePorter in One Command. Everything else is inside.
Ask your infrastructure a question. Get an answer backed by live data. A Self-hosted Conversational AI Agent capable of executing real Infrastructure Operations using Tools Ecosystem.
Control your tool ecosystem from a single sidebar, and interact using natural language – AppTelePorter orchestrates the execution
Meet Tool Studio – the built-in environment for crafting your custom AppTelePorter plugins with zero deployment overhead
Crrently Supported External LLMs and Workflow Tool Integrations
The Future of Infrastructure Operations
A featured platform for Infrastructure AI
AppTelePorter enables teams to execute infrastructure operations, automations, and workflows through simple conversations—while keeping credentials protected, approvals enforced, and every action fully under your control.
Deployment Options
Encrypted Credential Vault
Workflow & API Integration
Testimonials & Reviews
Hear From Our Happy Clients: Their Stories
Lorem ipsum is placeholder text commonly used in the graphic, print, and publishing industries for previewing layouts and visual mockups.
Emily R.
Marketing Lead
We switched from a legacy tool, and I can’t believe the difference. Setup was a breeze, and every question I had was answered within minutes. That level of care is rare.
Mark T
Startup Founder
I love how everything just works. No steep learning curve, no messy configurations—just a clean, powerful product that delivers what it promises.
John Doe
Founder
Beyond the robust features, what really impressed me was the onboarding support. They walked me through everything step-by-step and made us feel valued.
Jason M
Product Manager
The features are great, but the support team truly sets this SaaS apart. From onboarding to daily use, their quick responses and helpful guidance made the whole experience smooth and stress-free. It feels like we have a real partner, not just a service.
Lena K
UX Consultant
This SaaS solution is a gem. It adapts perfectly to our needs, scales effortlessly, and the team behind it feels more like a partner than a provider.
Aliana Lorel
Founder – Lorel Technology
As someone who’s tried countless tools, this one stands out. Performance is solid, updates are regular, and customer service? Simply outstanding.
Sara D
Freelancer
This platform completely transformed how we manage our workflow. The intuitive UI, along with a support team that truly cares, makes it one of the best SaaS experiences I’ve had.
David V
IT Director
Smooth, fast, and reliable. Our team instantly became more productive. The personalized help during onboarding made a huge difference.
Peter Brandson
Owner – Brandson Industry
User-friendly, beautifully designed, and packed with functionality. I’ve never experienced such responsive support. Worth every penny.
Frequently Asked Questions
Deploy with Absolute Confidence.
AppTelePorter combines an intuitive chat experience with powerful automation. Browse the answers below to learn about our built-in Tool Studio, encrypted credential vault, flexible LLM routing, and n8n integrations.
Does it require a GPU?
No, a GPU is not mandatory. AppTelePorter runs seamlessly on CPU machines, as the runtime base image has no GPU dependencies. However, if a GPU is available and detected at container startup, the system automatically uses it when running the inbuilt model in the Max version. While GPU inference is faster, the platform is fully functional on standard CPU hardware.
Does AppTelePorter work without an internet connection?
Yes, the platform is architected for local-first execution, though the extent of its offline capabilities depends on your deployed edition.
Max Edition (100% Air-Gapped): The AI model is bundled directly into the container image. After the initial docker pull, the entire system—including conversational chat via the local Ollama inference engine—operates completely offline on your hardware with zero outbound internet connectivity.
Lite Edition (Hybrid Offline): In this edition, an internet connection is required to route conversational inference to your chosen external LLM provider (OpenAI, Gemini, or Claude). However, the core platform remains fully localized. Critical functionalities including the Tool Studio, Device Beacon, your encrypted credential vault, and local API execution (such as local n8n integrations) are processed entirely on your machine and will remain available and functional even if external internet access drops.
Can I use an external LLM as the core intelligence engine in AppTelePorter?
Yes, absolutely. You can bypass the local model and route all inference processing directly to frontier cloud engines. Currently, AppTelePorter natively supports OpenAI (ChatGPT), Anthropic Claude, and Google Gemini, with support for additional models actively expanding in future versions. To configure your external provider, simply navigate to Settings → LLM Settings in the dashboard, select your provider, and securely paste your API token. All your credentials remain localized inside your container vault—only the context required for the immediate query synthesis pass is securely dispatched.
How are device credentials protected?
Security is foundational to AppTelePorter’s architecture, guaranteeing that your credentials are never exposed to external LLMs, chat histories, SSE payloads, or log files. The platform employs a strict, zero-knowledge encrypted vault that operates entirely on your local hardware.
Here is exactly how your infrastructure secrets are secured:
Advanced Cryptographic Stack: Your secrets are protected by a rigorous two-layer cryptographic design. Your master password is run through a highly secure, memory-hard key derivation function—to generate a 256-bit encryption key. This key encrypts the entire vault into a single blob using Fernet authenticated encryption
Zero Prompt Exposure: When the AI agent needs to execute an action, it references your credentials using abstracted tokens (e.g., $vault.name.field). The platform resolves these tokens into actual credentials in-memory immediately prior to tool execution. Your raw passwords are never serialised into an LLM prompt.
Strict No-Backdoor Policy: AppTelePorter never stores your master password on disk, in a database, or anywhere else. Only a bcrypt hash is kept securely inside the vault itself, used solely to verify your identity at unlock time.
Important Consequence: Because of this uncompromising zero-knowledge architecture, there are no password resets, no recovery emails, and absolutely no backdoors. If you lose your master password, the encrypted vault blob cannot be decrypted under any circumstances, and all stored credentials, API keys, and user accounts will need to be recreated from scratch
Can I add my own Python scripts to AppTelePorter and run them as AI-executable tools?
Yes, absolutely. AppTelePorter is built from the ground up to be fully extensible, allowing you to turn your custom Python automation scripts into production-ready tools that the AI can understand and execute.
You can build and deploy your automations directly through the platform using two flexible methods:
The Built-In Tool Studio: AppTelePorter features a browser-based IDE called Tool Studio. It comes pre-loaded with boilerplate templates and live examples. You can write your Python execution logic (handler.py) and define the AI parameter schema (tool.yaml) directly in your browser. The platform hot-reloads your code instantly without requiring a container restart.
Custom MCP Servers: You can write or import full Model Context Protocol (MCP) servers. AppTelePorter will automatically discover them, spin them up as isolated child processes, and expose their functionality directly to the chat interface
Can I integrate AppTelePorter with exsting automation pipelines like n8n?
Absolutely. The platform exposes a fully standard API surface (including endpoints like /v1/chat/completions and /v1/responses). This means any workflow, n8n pipeline, or custom HTTP client already built to target standard cloud LLMs can simply be repointed to your local AppTelePorter instance by updating the base URL. Because it utilizes the exact same Bearer token authorization format and request structure, you can transition your infrastructure to a private model with zero code changes.
Additionally, AppTelePorter’s execution engine isn’t limited to the chat interface. All registered tools and plugins can be accessed directly via dedicated REST endpoints. This allows you to seamlessly expose your AppTelePorter tools to other external LLM agents or automation platforms (like n8n HTTP Request nodes) as standard HTTP endpoints (e.g., POST /api/tools/{tool_name}/run). You can pass tool parameters directly through the API for immediate, headless execution across your network
How is the platform updated, and will I lose my custom plugins?
Updating is as simple as pulling the new Docker image and restarting the container with your existing volume mount. All your persistent state—including your encrypted vault, custom plugins, MCP servers, and session history—lives safely in your local appteleporter-data/ directory. Upgrades will never overwrite or touch your custom tools.
Start Building With AppTelePorter Today.
The Community Edition is free and will always remain free. Get started in minutes, connect existing tools, build your own automations, and explore AI-powered infrastructure operations. Create custom plugins and MCP servers, contribute to the ecosystem, and share your work with the community through the AppTelePorter GitHub repository.