Full-Stack Engineer + AI
We are expanding our internal geospatial analytics platform that combines time-series data with interactive map visualizations for Russian and international clients. The system delivers near real-time statistics, metrics monitoring, and configurable analytics. As a Full-Stack Engineer, you will design and build high-performance interfaces, develop backend systems, and integrate an AI layer to enhance automation and business outcomes.
Our stack (today):
- Frontend: React + TypeScript, contemporary tooling, real-time visualization (WebGL/maps), and solid CI-backed testing
- Backend: Go, Python (FastAPI/asyncio), ClickHouse, PostgreSQL, Redis, MongoDB, Kafka
- Infra: Docker, Kubernetes, Git-based CI/CD, cloud-agnostic deployments, Yandex internal infra
- Planned: MCP servers, (M)LLMs, vector stores, and CI/CD for containerized MCP
Tasks that await you
Migrating core backend services to Go
You will rewrite our Python backend to Go incrementally, building gRPC services with protobuf. Your focus will be on performance-critical paths like query execution, geospatial operations, and real-time aggregation. You will also establish proper patterns for observability, error handling, and testing.
You will rewrite our Python backend to Go incrementally, building gRPC services with protobuf. Your focus will be on performance-critical paths like query execution, geospatial operations, and real-time aggregation. You will also establish proper patterns for observability, error handling, and testing.
Introducing AI via MCP servers
You will design and ship MCP servers that safely expose Atlas tools (ClickHouse queries, geospatial operations, map snapshots, dashboard CRUD) with tool/resource definitions, multi-tenant RBAC, auditing, and rate limiting.
Building the “Ask Atlas” conversational UI
You will develop a chat dock with streaming responses, citations, and one-click actions that invoke MCP tools. By leveraging NL→SQL/Geo pipelines translating natural-language questions into ClickHouse SQL and deck.gl layer configs, you will enable features such as anomaly detection and alert triage by combining LLM reasoning with statistical detectors and MCP tool calls.
We expect that you
- Are proficient with React and TypeScript (hooks, state management, charting)
- Have experience with Go backend development, API design, and testing
- Are comfortable with SQL and NoSQL, including PostgreSQL
- Have experience with Linux development and distributed systems (microservices, observability, resilience)
- Are product-minded and eager to understand user problems deeply
- Have professional working knowledge of English and Russian
Preferred qualifications
- Have experience with Python backend development (FastAPI + asyncio), API design, and testing
- Have experience with ClickHouse and/or MongoDB
- Have worked with LLMs via MCP or function-calling, building chat UX in React
- Have built or integrated MCP servers, retrieval/RAG pipelines, or policy/guardrail frameworks
- Have experience with Kafka, event pipelines, or stream processing
- Know how to operate Kubernetes deployments in public clouds (AWS/GCP/Azure/Yandex Cloud)
- Have worked with geospatial frameworks (deck.gl, H3, quadkeys, GeoJSON, WebGL)
- Have an interest in product/technical metrics and proposing improvements
- Have experience with gRPC(-web) and protobuf
Benefits
1
Health
2
Growth
3
Sport
4
And more
Enhanced medical insurance begins from your first month at Yandex. It covers health care services at clinics and hospitals, whether for emergency care, surgeries, routine visits in your home country or abroad, medical tests and diagnostics, dental care, accident insurance, etc.
Additionally:
- Corporate insurance plans for family members (children and spouses) on special terms.
- Mental health programs.
At Yandex, we have everything you need for continuous growth and learning: an internal educational platform, mentoring, and programs for both beginners and experienced leaders.
We also cover the cost of attending industry conferences — both as a speaker and as a participant.
Moreover, at Yandex, we have internal projects where our employees share their expertise, discuss complex topics, and analyze their project cases.
At Yandex's big offices, we have our own fully equipped gyms. There, you can work out independently or with a trainer.
Additionally:
- Free online fitness classes.
- Additional discounts at fitness clubs and/or other fitness providers.
- Reimbursement of sports competitions.
- Corporate sport form for official sports competitions.
At Yandex, we promote health and fitness and have sports clubs and communities. You are welcome to join the existing teams or groups or create your own.
You can join existing teams or create your own.
Flexible schedule
We don't have a set start and end time for the workday — work when it's convenient for you and your team.
Hybrid format
Work where it's most convenient for you — for example, part of the week from the office and part from home.
Everything for children
Medical Insurance plans, kids' days at the office, gifts for newborns.
Wed Aug 07 2024 14:43:03 GMT+0300 (Moscow Standard Time)
