Middle AI Engineer

Our core values are innovation, quality, and customer satisfaction. We strive to deliver the best solutions and services, continuously seek new growth opportunities, and stay up to date with the latest trends and technologies in the industry.

This role focuses on integrating artificial intelligence into our production systems, including payment platforms, fraud prevention solutions, and customer interaction tools. The candidate is expected to solve real-world problems using large language models (LLMs), natural language processing (NLP), and machine learning techniques. Responsibilities include end-to-end development: building models from scratch, optimizing them, turning them into APIs, and integrating them with live systems.

This position requires an engineering mindset that goes beyond research to deliver production-grade solutions. Expectations include defining problems, constructing relevant datasets, training models, integrating outputs into products, and scaling systems sustainably. Every code component deployed to production must meet performance, security, and maintainability standards.

The role also involves developing custom LLM-based solutions such as document comparison, text segmentation, RAG-based search, and chatbot architectures, as well as integrating with vector databases (FAISS, Pinecone) and building multi-agent or rule-based logic systems.

Core Responsibilities:

• Design LLM-based systems (chatbots, summarizers, classifiers) and apply embedding-based semantic search;
• Rapid LLM prototyping using Ollama in local environments with low-latency testing;
• Build similarity-based retrieval systems using embeddings and vector databases (FAISS, Pinecone);
• Implement multi-step user interaction flows with Modular Command Processing (MCP);
• Enable LLMs to interact with external systems via toolcalling (e.g., databases, APIs, computation services);
• Use Python for data processing, model prototyping, and building production-level API services;
• Serve models via microservice architecture using FastAPI, with a focus on security and performance;
• Manage model I/O data using SQL and PostgreSQL, optimize query performance;
• Use Redis for caching, session management, and fast-access transient model outputs;
• Apply MLOps principles to manage pipelines, fine-tuning, versioning, and monitoring;
• Fully integrate developed systems into production, follow release and monitoring processes per MLOps standards;
• Align models with business goals and validate with A/B testing;
• Deliver projects end-to-end in direct collaboration with product managers, frontend/backend developers, and data analysts.

Technical Stack:

• Programming: Python (advanced), FastAPI;
• Databases: SQL, PostgreSQL, Redi;
• LLM & NLP: Hugging Face Transformers, LangChain, LlamaIndex, Toolcall, MCP;
• Embedding & RAG: Embedding generation and querying, FAISS, Pinecone, Milvus;
• Model Development: Ollama (mandatory), PyTorch, TensorFlow, scikit-learn (preferred);
• MLOps: Docker, Git (mandatory), MLflow, Airflow, Prefect (preferred);
• Security: Pydantic, cryptography, OAuth2, data validation.

Soft Skills:

• Clear engineering communication and ability to document technical decisions;
• Data-driven decision-making, testing, version control, and rollback management;
• Mandatory: Technical communication in Russian;
• Required: English at B2–C1 proficiency level.

We offer:

• Participation in exciting and large-scale projects;
• Challenging and engaging tasks that help develop your skills;
• Competitive salary, reviewed annually, along with a full social package;
• Comfortable work environment with an excellent work-life balance;
• Stable work schedule (Monday - Friday, 09:00–18:00), with a one-hour lunch break;
• Free English language courses;
• Opportunities for personal and professional development, including additional courses;
• Team-building activities, corporate gifts, and bonuses for special occasions (e.g., birthday, childbirth);
• Cozy office located near Valea Trandafirilor Park, at Decebal 6 (Botanica);
• Coffee and snacks at the office to keep you energized and in a good mood.

To apply for this job, candidates should send their CV to hr.office@paynet.md with subject line: „Middle AI Engineer”.