AI / Machine Learning Engineer

Quick Facts

• Role: AI / Machine Learning Engineer

• Focus: Generative AI/LLM solutions for ERP automation and a SaaS knowledge base + AI agent

• Key themes: RAG, vector databases, agent frameworks, continual learning, data/ML pipelines in the cloud

Description

You will support two innovative technology projects: AI assistance for ERP systems to automate accounting document processing, and a dynamic knowledge base with an AI agent for a SaaS application. The work includes building and deploying LLM-based classification and support workflows, integrating a conversational interface, and continuously improving the system from ongoing interactions.

Responsibilities

• Design and implement AI mechanisms for automatic accounting document settlement and booking

• Classify costs using LLMs based on invoices and counterparty data

• Build self-learning mechanisms to improve classification accuracy over time

• Scale the solution to 10,000+ instances of cloud accounting

• Implement RAG with vector databases for deep analysis of historical documents

• Integrate a conversational AI interface (chatbot/agent) with service tools (2nd-line support)

• Use LLMs to interpret, group, and auto-summarize user tickets and generate solution recommendations

• Continuously update training of the knowledge base based on current cloud interactions

Requirements

• Very good Python knowledge

• Proficiency with PyTorch and TensorFlow/Keras

• Experience with scikit-learn, Hugging Face, NumPy, pandas

• Ability to select, compare, and fine-tune LLM/OLM models (e.g., OpenAI, Claude, Mistral, Llama)

• Knowledge of Transformer architecture, prompt engineering, and fine-tuning (LoRA)

• Ability to evaluate model quality using precision, recall, F1, perplexity

• Experience building and maintaining vector databases and embedding pipelines

• Hands-on with FAISS, Pinecone, Milvus, Weaviate, or Chroma

• Experience with AI agent frameworks: LangChain, AutoGen, Haystack, LlamaIndex

• Knowledge of Model Context Protocol (MCP)

• Experience designing architectures with continual learning and running data pipelines with Airflow/Prefect/Dagster

• Experience with event-driven systems (Kafka, Redis Streams)

• Cloud experience: AWS, GCP, or Azure

• Experience cleaning and augmenting datasets

• Solid SQL skills

• Ability to work with unstructured data (text, images, audio)

• Polish language proficiency at least B2

Benefits

• Interesting projects matched to your skills and interests

• B2B contract without exclusivity requirements

• Career growth through work with the latest technologies across diverse projects