Deploy & Configure Bolt.New with Custom AI
Budget: €30 – €250 EUR
Scope:
1. Ensure Bolt.New is fully working on the server (AlmaLinux 8, Docker-based).
2. Use my subdomain, with working SSL (Let’s Encrypt / certbot preferred).
3. Replace the original AI backend (Ollama) with a self-hosted alternative, preferably:
• Mistral 7B Instruct
• Running via Docker (e.g., llama-cpp-python or text-generation-webui)
• Exposed via a REST API (/generate) that accepts prompts and returns completions
4. The AI will only need to support code generation (HTML, JS, CSS chunks)
Notes:
• You will receive SSH access on a restricted port with a jailed user.
• Docker and Docker Compose are installed.
Skills Required:
• Docker / Docker Compose
• Reverse proxy (nginx or traefik)
• Linux (AlmaLinux / CentOS)
• AI model integration (Mistral / LLMs / REST APIs)
• Bonus: Experience with Bolt.New or frontend builders
1. Ensure Bolt.New is fully working on the server (AlmaLinux 8, Docker-based).
2. Use my subdomain, with working SSL (Let’s Encrypt / certbot preferred).
3. Replace the original AI backend (Ollama) with a self-hosted alternative, preferably:
• Mistral 7B Instruct
• Running via Docker (e.g., llama-cpp-python or text-generation-webui)
• Exposed via a REST API (/generate) that accepts prompts and returns completions
4. The AI will only need to support code generation (HTML, JS, CSS chunks)
Notes:
• You will receive SSH access on a restricted port with a jailed user.
• Docker and Docker Compose are installed.
Skills Required:
• Docker / Docker Compose
• Reverse proxy (nginx or traefik)
• Linux (AlmaLinux / CentOS)
• AI model integration (Mistral / LLMs / REST APIs)
• Bonus: Experience with Bolt.New or frontend builders