Documentation: Claude API with Ollama embeddings + Railway deployment guide

Description

Add comprehensive documentation on how to use Claude API alongside Ollama embedding models, plus a Railway deployment guide with Ollama support.

Files to Look At

Documentation

• docs/self-hosting/docker.mdx - Current Docker deployment docs (Railway button already added)
• docs/self-hosting/overview.mdx - Self-hosting overview
• docs/self-hosting/environment-variables.mdx - Environment variable configuration
• docs/providers/claude.mdx - Claude provider documentation
• README.md - Main README with deployment options

Provider Configuration

• apps/webapp/app/lib/batch/providers/anthropic.ts - Anthropic/Claude provider implementation
• apps/webapp/app/lib/batch/providers/openai.ts - OpenAI provider for reference

Tasks

• Document how to configure Claude API key for chat/inference
• Document how to configure Ollama for embeddings only
• Add example environment variables for hybrid setup
• Create Railway deployment guide with Ollama service configuration
• Add troubleshooting section for common issues

Guidelines

1. The Railway deployment template may need updates to include Ollama service
2. Check existing environment variable patterns in environment-variables.mdx
3. Follow the documentation style in existing provider docs
4. Consider adding a diagram showing the hybrid architecture

Labels

good first issue, documentation

Read docs/self-hosting/docker.mdx, overview.mdx, environment-variables.mdx, and docs/providers/claude.mdx first, then inspect the Anthropic and OpenAI provider implementations for configuration patterns. Done means the docs explain Claude chat/inference, Ollama embeddings, hybrid environment variables, Railway deployment, and troubleshooting, with any required template updates and an optional architecture diagram covered.