Every time you ask ChatGPT a question, your data travels through servers you do not control, in a country whose privacy laws may not match yours. For many teams, that is a deal-breaker.

Self hosted AI changes the equation entirely. When you run your own AI stack, every prompt, every document, and every answer stays inside your network. No third-party processors, no cross-border data transfers, no surprise price hikes.

Synaplan is a fully open-source AI platform designed to be self hosted from day one. It gives you a single management layer for multiple AI providers — running on your own servers, under your own rules.

Why Self Hosted AI Matters More Than Ever

The regulatory landscape has caught up with the AI hype. The EU AI Act is now in effect, GDPR enforcement exceeded EUR 4.5 billion in fines by early 2026, and sector-specific rules like HIPAA, PCI DSS, and DORA add further pressure. Cloud-based AI services force you to navigate Data Processing Agreements, cross-border transfer assessments, and vendor audits — all of which disappear when you self host.

But compliance is only part of the story. The economics have shifted too. Open-weight models from Meta, Mistral, and DeepSeek now match proprietary cloud models for most enterprise tasks. Self hosting can be up to 18x cheaper per million tokens than premium cloud APIs for high-volume workloads.

What Makes Synaplan Different

Synaplan is not another chatbot wrapper. It is a knowledge management platform that connects your team to AI through a unified interface — all while keeping your data on-premise.

Multi-Provider Flexibility

Instead of locking into a single vendor, Synaplan lets you route requests to OpenAI, Anthropic, Google Gemini, Groq, Mistral, or local models via Ollama — all from the same chat interface. Switch models per conversation or per team. If a provider raises prices or changes terms, you swap it out without touching your integration code.

Real-Time Web Search with BRAVE API

Here is where things get interesting. Self hosted AI does not have to mean isolated AI.

Synaplan integrates a real-time web search powered by the BRAVE Search API, giving your AI access to current information from across the internet — without sending your private data to a search engine.

When a user asks a question that requires up-to-date information, Synaplan automatically searches the web, retrieves sources, and weaves the results into its response — complete with numbered citations and clickable source cards. The AI knows what it found online versus what it already knew, and it shows the difference transparently.

This means your self hosted AI is never out of date. Whether your team needs the latest regulatory changes, competitor pricing, or breaking news, the answer comes back grounded in real sources — all processed on your own infrastructure.

Memories: Your AI Learns and Remembers

Most AI tools start fresh with every conversation. Synaplan is different.

The Memories feature uses vector search (MariaDB VECTOR + Qdrant) to build a persistent knowledge layer. Every important fact, preference, or context your team shares gets stored and automatically retrieved in future conversations. Your AI remembers project details, client preferences, internal terminology — without you having to repeat yourself.

This is Retrieval-Augmented Generation (RAG) done right: your documents and knowledge base are indexed locally, and the AI pulls relevant context before generating any response.

Embeddable Chat Widgets

Need AI on your website or internal portal? Synaplan ships with a fully customisable chat widget that works cross-origin and detects its API URL at runtime. Embed it on your support page, intranet, or customer portal — all traffic routes back to your self hosted instance.

Multi-Channel Integration

Beyond the web interface, Synaplan connects to WhatsApp and Email, so your team (or your customers) can interact with AI through the channels they already use. Every message flows through your servers, with full audit trails.

The Technical Stack

Synaplan runs on a proven, modern stack:

Backend: PHP 8.3 with Symfony 7, served by FrankenPHP (Caddy-based)
Frontend: Vue 3 with TypeScript and Vite
Infrastructure: Docker Compose — one command to deploy
Database: MariaDB with vector extensions + Qdrant for semantic search
AI Providers: Ollama (local), OpenAI, Anthropic, Google, Groq, Mistral

Everything is containerised. A docker compose up -d gets you a running instance. The entire source code is on GitHub under the Apache 2.0 licence — fork it, audit it, extend it.

Self Hosted AI in Practice: What Your Setup Looks Like

Deploying Synaplan on your own servers takes minutes, not months:

Clone the repository from GitHub
Configure your .env with your preferred AI providers and API keys
Run docker compose up -d — the full stack starts, including the database, vector search engine, and web server
Upload your documents for RAG — Synaplan indexes them automatically
Enable web search by adding your BRAVE API key
Embed the chat widget on your website or share the internal URL with your team

No cloud account needed. No per-seat licensing. No data leaving your premises.

Who Benefits Most from Self Hosted AI

Regulated industries (finance, healthcare, legal) where data residency is non-negotiable
Enterprises tired of unpredictable API costs and vendor lock-in
Government agencies that must keep data within national borders
Privacy-conscious organisations that want AI benefits without the surveillance trade-off
Development teams that need to inspect, modify, and extend their AI tooling

The Cost Advantage

For organisations processing more than a few million tokens daily, self hosted AI pays for itself quickly. No per-request charges, no token-based pricing — just your server costs. Synaplan is open source, so there are no licence fees on top.

Compare that to cloud AI services where costs scale linearly with usage and a single busy month can blow through your budget.

Getting Started

Synaplan is free, open source, and ready to deploy:

GitHub: github.com/metadist/synaplan
Website: www.synaplan.com

Self hosted AI is no longer a luxury reserved for companies with dedicated ML teams. With Synaplan, any organisation can run a production-grade AI platform on their own infrastructure — complete with web search, persistent memory, multi-provider routing, and full data sovereignty.

Your data. Your servers. Your AI.