By agreement
Moscow · 3-5 years · 5/2 · Permanent contract
The AI Platform Lead is responsible for building and developing a corporate AI platform for LLM and DS models from scratch: a unified point for training, fine-tuning, inference of models, working with knowledge, and creating AI agents on a hybrid infrastructure (on-prem GPU and cloud). A separate area of responsibility is defining the architecture and standards for how product teams build AI products and AI agents on the platform.
Responsibilities:
- Developing the strategy, roadmap, and target architecture for the AI platform, synchronizing with CDO goals and business priorities.
- Developing and scaling the team: structure, hiring, onboarding, development, and retention of engineers (ML/LLM, platform, MLOps).
- Defining the architecture of the AI platform and AI agents: reference architectures, patterns, component selection (in-house development, open source, vendors), conducting PoCs.
- Developing and implementing standards for product teams: how to build AI products and AI agents on the platform; reviewing architectural solutions, documentation, training, and competency centers.
- Managing the model lifecycle: training, fine-tuning, versioning, deployment, monitoring, evaluation, and decommissioning.
- Developing MLOps/LLMOps tools: model CI/CD, experiments, model registry, feature store, observability, and data quality for ML.
- Managing inference: hybrid serving (on-prem GPU and cloud), SLAs, quotas, workload prioritization, performance, and cost optimization.
- Developing RAG and knowledge work: pipelines, vector stores, search quality, and relevance.
- Developing the agent platform: tools, memory, sandbox, guardrails, security, and observability of AI agents.
- Working with internal customers – product teams and the AI SDLC group: gathering requirements, consultations, prioritizing the platform backlog; developing the platform as a product (DevEx, self-service portal, documentation).
- Ensuring security and compliance: AI usage policy, data and personal data protection (152-FZ), auditing, traceability of model and agent solutions.
- Managing the budget of the division and reporting to the CDO on metrics: inference SLA, inference cost/token, model onboarding time, adoption.
- Technology radar: monitoring the AI tools and vendors market, evaluating open-source solutions, piloting, and implementation.
What we expect:
- Experience in building and developing an internal ML/LLM platform in a large company: from architecture and component selection to production launch and scaling.
- Practical industry experience with LLMs: inference and serving (vLLM or similar), fine-tuning, RAG, model evaluation, performance optimization (quantization, batching).
- Experience with the classic ML/DS stack: training and deploying models, MLOps (model registry, versioning, CI/CD, monitoring), feature store, data pipelines.
- Experience with hybrid infrastructure: on-prem GPU clusters and cloud LLM services, Kubernetes, GPU resource planning, cost management (FinOps).
- Experience in designing AI agents and agent architectures: orchestration, tool calling, guardrails, understanding agent building patterns.
- Experience in developing and implementing standards and architectural patterns, architecture review, working with internal customers; building a platform as a product (DevEx, documentation, self-service).
- Experience in managing and developing an engineering team: hiring, developing, and retaining employees, scaling the team, establishing processes.
- Developed management and communication skills: working with top management, translating business tasks into platform solutions, ability to explain the value of AI to business.
- Knowledge of the AI tools market and experience in selecting solutions: in-house development vs. open source vs. purchase, evaluation, and integration.
- Understanding of AI security and compliance: personal data protection (152-FZ), guardrails, responsible AI use, auditing, and traceability.
Will be a plus:
- Experience in retail / e-commerce / DIY.
- Experience launching an AI platform from scratch (greenfield) in a large company.
- Experience integrating LLMs into the SDLC and creating developer assistants (AI-assisted coding, code review).
- Experience with domestic LLMs and clouds (Yandex Cloud, GigaChat/YandexGPT) and with foreign models.
- Experience in building a Data/Feature platform and interacting with the data platform team.
- English language for reading documentation and communicating with vendors.
Working with us means:
- Working in a company with a developed engineering culture.
- Flexible bonus system.
- Extended social package: voluntary medical insurance with dentistry from the first month of employment, psychologist and life insurance, meal and mobile communication compensation.
- Opportunities for professional growth, development programs for employees.
- Corporate training and access to a knowledge base.
- Internal professional communities and events.
- Work autonomy, the ability to change rules, make mistakes, and create new things.
- Work format - office.
- Modern office a 2-minute walk from MCC ZIL.
Every year, more additional options become available at the employee's choice. For example, partial compensation for travel, training expenses, sports activities.
Work location
Moscow
Likhachev Prospect, 15
LM Tech
