About the Company and Direction

PSM Unlim is developing a local AI infrastructure and applied AI projects. We are creating an AIBase / AICluster / AIFactory framework: engineering base, computing, model layer, RAG, OCR, AIGateway, integrations, information security, and operation of AI services within a closed environment.

We need a senior engineer who will become the technical core of AIFactoryPSMUnlim: help design and build a local AI platform, as well as lead junior specialists, train them, and form engineering standards for the direction.

What You Will Do

• design and develop the backend of a local AI platform;
• create AIGateway / Router for request routing;
• develop ModelRegistry and model version management;
• build a RAG pipeline: ingestion, OCR, parsing, chunking, embeddings, retrieval;
• integrate Qdrant / Milvus / similar solutions;
• connect local and external LLM / VLM;
• design APIs for AI services;
• integrate the AI pipeline with 1C / ERP / Bitrix24 / CRM / DMS / Service Desk / MES / SCADA;
• build monitoring for response quality, latency, errors, logs, and load;
• participate in client pilots and technical solutions for AICluster / AIFactory.
• form internal standards for the development and operation of the AI platform: service templates, documentation, test sets, logging, and metrics.
• lead junior specialists: assign tasks, conduct code reviews, train in RAG / OCR / LLMOps / backend approaches;

What's Important

• production backend is more important than academic ML;
• experience with RAG and integrations is more important than "tried ChatGPTAPI";
• experience with local models, GPU inference, and closed environments is a strong plus;
• the ability to work with information security, DevOps, architects, and business stakeholders is mandatory;
• we need someone who can be a technical leader and mentor for the team, not just a task executor.
• experience in mentoring, code review, task assignment, and training junior / middle specialists is a mandatory part of the role.

Tech Stack

Python, FastAPI, Docker, Kubernetes, PostgreSQL, Redis, Kafka / RabbitMQ, Qdrant / Milvus, LangChain / LlamaIndex or custom RAG architecture, vLLM / SGLang / Ollama / TGI as a plus, MLflow / ClearML / DVC as a plus, Prometheus / Grafana, GitLabCI/CD, OCR / parsing, LLM / RAG, GPU inference.

What We Offer

• our employees receive exclusively white salary: (base salary + bonus + annual bonus). Figures will be discussed individually with each candidate.
• we are constantly developing and growing: training funded by the company, TOP trainers and speakers, participation in seminars and professional conferences;
• the opportunity to build a high-level industrial local AI platform from scratch;
• the role of the technical core of AIFactoryPSMUnlim; influence on architecture, tech stack, team, and product strategy;
• the opportunity to form the AIPlatform team: train junior specialists, set standards, and grow future engineers for the direction;