AI / ML Engineer

About Momentum:

At Momentum, we accelerate digital innovation in healthcare. Our mission is to empower HealthTech startups and scaleups by transforming ideas into cutting-edge digital products. We combine deep industry expertise, strategic consulting, and full-stack development to create solutions that drive growth, compliance, and patient impact.

We believe in technology that makes healthcare more accessible, efficient, and human-centere

About the project

‍An education-sector client is modernizing its existing platform and building new AI-based solutions. We are recruiting several people across two independent roles described below.

Scope

‍A newly formed team designing and shipping AI-based solutions: language models (LLMs), vector databases, and RAG architectures. Example team projects:

• AI support in an ERP system: automatic document posting and cost classification based on invoice content and line items, a self-improving mechanism that learns from accounting actions, RAG and vector databases over historical documents, at a scale of 10,000+ cloud instances.
• Dynamic knowledge base for SaaS users: a base built and continuously updated from second-line support, LLM-driven grouping of tickets and summary generation, accessed through a conversational agent integrated with client applications.

Requirements

• Python: strong command, clean and modular production code.
• Hands-on with ML/AI libraries: PyTorch, TensorFlow/Keras, scikit-learn, Hugging Face; data handling (NumPy, pandas); integrating models with external APIs.
• Working with LLMs (OpenAI, Claude, Mistral, Llama): selecting, comparing, and prompting; understanding Transformer architectures, tokenization, and context management; awareness of model quality metrics (precision, recall, F1, perplexity).
• RAG in practice: vector databases (FAISS, Pinecone, Milvus, Weaviate, Chroma), semantic search, embedding management, and integrating them with LLMs.
• Building and shipping AI features in the cloud (AWS, GCP, Azure).
• Polish at C1 level (communication with the client team).

Nice to have

• Fine-tuning / LoRA and adapting NLP models.
• Agentic architectures and MCP: frameworks (LangChain, AutoGen, Haystack, LlamaIndex), AI agents that use tools and APIs, and the Model Context Protocol.
• Designing end-to-end AI architecture with continuous background ingestion into the vector store.
• Data pipelines (Airflow, Prefect, Dagster) and event-driven architecture (Kafka, Redis Streams); data and model versioning, continual learning.
• TypeScript, Node.js, or Go for backend integrations.
• Data work: dataset preparation, cleaning, and augmentation; SQL and big data tools; unstructured data (text, image, audio).
• English at B2 level.

Recruitment process

‍A short, focused process, the same for both roles:

1. Screening and technical interview on Momentum'w side.
2. Technical interview with the client.

Questions? Get in touch.

‍