Principal ML Researcher: Remote Post-Training & RL Lead

Toloka; Sevilla; Remote; EUR 150,000 - 210,000; Full time

Benefits offered by this job

Remote or hybrid model

ESOP

IT setup allowance

Paid PTO

Job summary

Toloka AI is seeking a Principal ML Researcher to define the overarching technical vision and architecture for its core ML stack, bridging frontier AI research and platform engineering. You will lead multi-quarter strategy, mentor senior researchers, and collaborate with leadership to drive scalable, cost-aware ML solutions for real-world annotation workloads.

We offer a full remote or hybrid model, competitive compensation, ESOP, and IT setup allowances.

Qualifications

• 6+ years in ML engineering or applied research with leadership in post-training/LLaM alignment.
• Strong theoretical and practical mastery of LLM alignment methods (SFT/LoRA, RLHF/RLAIF, GRPO, DPO, PPO).
• Proven ability to build and maintain evaluation pipelines and calibrate LLMs against human standards.
• Expertise in distributed training systems (Python, PyTorch) and efficient inference (vLLM, TensorRT-LLM).
• Architectural vision to map complex problems into robust platform features balancing cost and reliability.
• Technical leadership with mentoring experience and cross-functional influence.
• Excellent written and spoken English for internal/external communication.

Responsibilities

• Drive multi-quarter strategy and architecture for Toloka's post-training stack.
• Lead greenfield post-training and RL pipelines (GRPO, PPO/DPO, reward modeling, RLAIF).
• Design automated evaluation ecosystems including LLM-as-a-judge concepts.
• Lead distillation, quantization, context compression, and latency-cost optimization.
• Shape agentic platform capabilities and tool-use workflows for multi-step reasoning.
• Provide technical leadership, reviews, and mentorship across ML researchers and engineers.
• Collaborate with Product/Engineering to translate client challenges into scalable architecture.

Skills

ML engineering

Leadership

LLM alignment

Python

PyTorch

Distributed training

Evaluation pipelines

Communication

Tools

Deepspeed

Megatron

FSDP

vLLM

TensorRT-LLM

Job description

Toloka AI is seeking a Principal ML Researcher to define the overarching technical vision and architecture for its core ML stack, bridging frontier AI research and platform engineering. You will lead multi-quarter strategy, mentor senior researchers, and collaborate with leadership to drive scalable, cost-aware ML solutions for real-world annotation workloads.

We offer a full remote or hybrid model, competitive compensation, ESOP, and IT setup allowances.