Senior AI Engineer (LLM)

O projekcie: How this role works: You will start by building the system as part of our team, working in a focused, greenfield setup. After a few months of introduction to the project, you will seamlessly continue working on the same product directly with the client, becoming part of the client’s team responsible for its long-term development and scaling. There is no handover phase and no context switching - you stay with the same codebase, product, and technical challenges, while gaining long-term ownership and impact. Biggest challenges in this role: - Greenfield project built from scratch - High-scale, user-facing systems with strict performance and reliability requirements - Designing systems meant for long-term ownership, not short-term delivery - Balancing model quality, latency, and cost in production LLM systems What you’ll learn: - How to build LLM-powered products from scratch and take them to production - Proven approaches to running LLMs in production at scale - How to design, evaluate, and evolve NLP systems used by real users - Best practices for production ML and AI system architecture What you’ll get to try and experiment with: - End-to-end ownership of LLM-based systems - Optimizing retrieval models, RAG pipelines, and inference workflows - Experimenting with different LLMs, prompting strategies, and system designs - Solving performance and reliability challenges under heavy traffic We want to offer you: - Work with an experienced team that continually shares knowledge and is not afraid of testing new solutions - Flexible hours - Individual work tools – Macbook Pro, Dell screen, JBL headphones? You can tailor the equipment to your needs - Sport & wellness benefit (Kafeteria MyBenefit) - Private medical care - A comprehensive benefits package after transitioning to our partner, including employee benefits, training opportunities, and occasional bonuses Wymagania: - At least 4 years of experience in ML / AI engineering - Proven experience owning production NLP or LLM systems - Strong understanding of Transformer architecture and deep learning foundations behind modern LLMs (attention mechanisms, self-attention, positional encoding, encoder–decoder vs decoder-only models) - Proven experience designing, fine-tuning, and deploying end-to-end NLP and deep learning solutions in production environments - Hands-on experience building LLM-powered production systems (e.g. GPT, Claude, Gemini), including prompt engineering, evaluation, supervised fine-tuning (SFT), instruction tuning, and parameter-efficient fine-tuning methods (e.g. LoRA, adapters), as well as user-facing integrations - Experience training and fine-tuning Transformer-based models using frameworks such as PyTorch and Hugging Face (including dataset preparation, tokenization strategies, hyperparameter tuning, and evaluation pipelines) - Strong understanding of scalability, performance optimization, and system design for large-scale ML/LLM systems - Python proficiency with experience building and maintaining reliable production services and data pipelines - Strong software engineering mindset, including code quality, testing, scalability, and production deployments - Experience building RESTful APIs (FastAPI, Flask) to expose ML/LLM capabilities - Curiosity and commitment to continuous learning in the NLP/LLM/AI space - Collaborative team-player with strong communication skills Codzienne zadania: - Designing, implementing, and deploying end-to-end NLP and deep learning systems - Building LLM-powered applications that interact with real users - Developing and maintaining production Python services - Exposing models and pipelines via REST APIs (FastAPI, Flask) - Working on retrieval models and techniques (RAG, embeddings, ranking) - Evaluating, monitoring, and continuously improving model and system quality - Scaling systems to handle enormous volumes of requests