Freelance / Contract | Remote | Start ASAP
Sunday Market Media is seeking a sharp AI workflow engineer or LLM modeling expert with strong n8n experience to help us debug, optimize, and scale our vector-enhanced automation flows.
We run a self-hosted n8n instance integrated with:
• PGVector (PostgreSQL vector storage)
• LLMs (OpenAI + Local models like Mistral)
• Dynamic embeddings + RAG querying
• Internal business knowledge base enrichment
🔧 Responsibilities
• Debug and optimize n8n workflows involving AI agents and vector memory
• Improve semantic search / term lookups via PGVector
• Ensure accurate and performant LLM responses across time-aware flows
• Help configure system prompts, embeddings, and metadata mapping
✅ Requirements
• Solid understanding of LLMs, embeddings, and retrieval workflows
• Experience with n8n, including Function nodes, HTTP/AI Agents, and Postgres integrations
• Proficiency with PGVector and managing vector data
• Comfortable with prompt engineering and AI response debugging
💡 Bonus Points
• Experience fine-tuning or self-hosting models (Mistral, GPT-J, etc.)
• Familiarity with embedding APIs (OpenAI, Hugging Face)
• Knowledge of vector hygiene, dimensionality mismatches, and memory management
