LanGPT

LanGPT is an AI assistant for academic tasks, integrated into the EBS Lan (Russian academic e-library platform). It helps students, researchers, and instructors with academic writing tasks: formulating hypotheses, choosing methodology, summarizing papers, drafting sections with citations, and checking requirements for academic content.

The RAG architecture answers from verified sources in the EBS Lan catalog, available to the company through partner agreements with publishers — without reaching out to the external internet.

Area of responsibility — backend and operations

I own the backend and infrastructure side of the product: implementing services, taking part in the Kubernetes deployment, and providing operational support.

• Public API: JWT sessions, rate limiting by pricing plan, SSE streaming of responses, and task queuing via RabbitMQ.
• RAG service over Elasticsearch for verified educational and academic content.
• The AI processing pipeline covers more than ten academic-writing tools, plus routing and prompt selection per task.
• Content moderation of user requests to keep undesirable and unsafe content out of an academic product.
• LLM backend balancer with a dynamic server registry, least-loaded / least-connections selection with priority, fallback, and OpenAI-compatible adapters.
• Production observability: metrics in Prometheus, dashboards in Grafana, structured logs and analytics in Elasticsearch.

Scale

The service is being piloted at 20 Russian universities; per the company's public materials, more than 37,000 requests and 6,500 active users during the testing period; the company's public plan for 2027 is 100 universities.

The figures come from the company's public materials about the product. The role and area of responsibility can be confirmed by the employer on official request.

Stack: Python · FastAPI · PostgreSQL · MySQL · RabbitMQ · Redis · Elasticsearch · vLLM · OpenAI-compatible APIs · Prometheus · Grafana · Docker · Kubernetes · Helm · GitLab CI