Freelance LLM Engineer — Integrate AI into Production

Monterail; Polska; Remote; PLN 165,000 - 331,000; Part time

Job summary

Monterail is seeking an experienced freelance LLM Engineer to design, build, and integrate AI features into existing Python, Node.js, or Ruby projects. The role focuses on delivery, collaborating with clients to implement practical AI solutions and ensure production readiness.

You will work on chat, summarization, classification, and smart search features, building efficient RAG pipelines with embeddings and vector databases, while considering cost, latency, and reliability.

Qualifications

• Strong software engineering background in Python, Node.js, or Ruby.
• Hands-on experience integrating LLM APIs into production systems.
• Experience building RAG pipelines, embeddings, and vector search.
• English B2/C1.
• Availability part-time or full-time (B2B contract).

Responsibilities

• Add AI functionality into existing Python, Node.js, and Ruby codebases.
• Build LLM-powered features: chat, summaries, classification, smart search, document Q&A.
• Design lightweight RAG pipelines using embeddings and vector search.
• Work with vector DBs (pgvector, Pinecone, Qdrant).
• Implement safe, reliable LLM endpoints (OpenAI, Anthropic, Azure).
• Work directly with clients to shape AI features and reduce manual effort.
• Advise clients when NOT to use AI and navigate trade-offs around latency, accuracy, and cost.

Skills

Python

Node.js

Ruby

LLM APIs

RAG pipelines

Embeddings

Vector search

Cloud/AI integration

Tools

OpenAI API

Anthropic API

Azure AI

pgvector

Pinecone

Qdrant

Job description

Monterail is seeking an experienced freelance LLM Engineer to design, build, and integrate AI features into existing Python, Node.js, or Ruby projects. The role focuses on delivery, collaborating with clients to implement practical AI solutions and ensure production readiness.

You will work on chat, summarization, classification, and smart search features, building efficient RAG pipelines with embeddings and vector databases, while considering cost, latency, and reliability.