NaPopravku is one of the leaders in Russian digital medicine. Our mission is to make illness less stressful in a person's life. We achieve this by implementing services that make medicine simpler and more convenient. It is important to us that everyone can take advantage of all the benefits of digital medicine using our service.

Our main product is finding a good doctor and convenient real-time appointment booking. Also, 3 years ago we created our own telemedicine service "NaPopravku Plus" – a subscription for online consultations with doctors, and in August 2025 we launched "Health Tracker" – created with doctors, enhanced with AI technologies.

We are 12 years old, and there are already over 300 of us, but in our hearts, we are a startup. We value people's ability to take responsibility and achieve results.

Among our partners are the largest private clinics and state medical centers, and most medical information systems.

We are looking for an AI Engineer to help develop and implement AI solutions into NaPopravku products.

In this role, you will work with LLMs, RAG, and AI agents: integrate models into the product, develop existing AI features, create internal AI tools, and help teams use new technologies more effectively.

What you will do:

• Develop health analysis systems: our key focus is AI that helps users understand their condition and see a doctor in time;
• Implement AI features into the product and the company: LLM integration (online and on-prem), intelligent search, assistants for users and internal teams;
• Develop existing AI solutions: AI Doctor and Health Tracker are products that are already working and that you will grow;
• Optimize internal company processes using AI: a separate area of tasks and projects is AI tools for internal teams;
• Participate in the development of Agents: pipeline orchestration, tools for agents, working with external APIs — you won't have to start from scratch with the team;
• Develop RAG/knowledge solutions: indexing pipelines, embeddings, vector databases, data updates, response quality control;
• Monitor quality: metrics and evaluation of prompts and models, conducting experiments.

Tech Stack:

• PHP 8.3–8.4 (Laravel), xDebug;
• Python 3.10+, notebook (Jupyter), LLM/embedding libraries;
• Postgres, ClickHouse, Redis; for RAG we use pgvector / Qdrant;
• Docker + Docker Swarm;
• Model providers and runtimes: OpenAI/Anthropic/DeepSeek, locally — Ollama/vLLM.

What we expect:

MUST (we won't hire without this).

• You have commercial development experience of 2+ years (we consider Middle grade);
• You regularly use AI agents (Claude Code, Codex) for work, constantly improving your work environment and tools;
• You are fluent in PHP, Laravel, REST, Git;
• You have real Python experience — not necessarily deep: our main backend is in PHP, AI products are in Python;
• Basic practical experience working with LLMs: calling models, understanding prompt-design principles, contextualization, working with tokenization and limitations;
• You understand the basics of RAG (when it's needed, what it consists of), you can critically evaluate the quality of model responses;
• You understand LLM risks: hallucinations, the need for response validation, model responsibility boundaries;
• You learn quickly, actively use ready-made solutions and services to avoid reinventing the wheel;
• You can plan your work based on the technical specification and provide adequate estimates.

Desirable (will be an additional plus).

• Experience with notebook environments (Jupyter/VS Code notebooks) for quick hypothesis testing;
• Practice with LangChain / LangGraph or similar (LlamaIndex, etc.) for building chains and agents;
• Working with vector databases (pgvector, Qdrant, Milvus) and embeddings;
• Knowledge of the ecosystem of providers/runtimes: OpenAI/Anthropic/Google, local runtimes ( Ollama, vLLM), Hugging Face;
• Experience in building quality monitoring of LLM features (logging, metrics, alerts);
• Experience developing in other languages: Go, JavaScript/TypeScript.

What we offer:

• Everything is transparent here: official employment and "white" salary (and we are an accredited IT company);
• Flexible start of the day (from 08:00 to 11:00 Moscow time), convenient work format for you (hybrid (St. Petersburg) / remote, 5/2 schedule;
• Smooth onboarding and support — no one will abandon you, we will help you integrate into the team and understand the processes;
• A team that is comfortable to work with — we share experience, support each other, and respect opinions;
• We have a grade system — a clear path for development and growth in the team;
• Health care: free access to "NaPopravku Plus" and Health Tracker for you and your family;
• BestBenefits — corporate discount program for employees: access to favorable offers from partners (purchases, services, travel, etc.);
• 3 Day Off — bonus to care for work-life balance;
• Necessary equipment and software;
• Spacious bright office near Vasileostrovskaya and Sportivnaya metro stations (within walking distance);
• PlayStation, board games, corporate library;
• Holidays and corporate events (can't do without joint recreation);
• Gifts from the company for Birthdays and other events.