Joi AI is an AI companion platform developing emotionally intelligent AI models. We are looking for a Lead ML Engineer to optimize LLMs and provide technical leadership to NLP and CV teams.
Responsibilities:
- Optimize and scale LLM inference (SGLang, vLLM, TensorRT-LLM).
- Work with models up to 1T+ parameters and distributed GPU infrastructure.
- Train and fine-tune LLMs, develop AI agents.
- Provide technical leadership to ML engineers and define technical strategy.
Requirements:
- Production experience in LLM inference optimization.
- Knowledge of distributed computing, MoE, KV-cache, quantization, and GPU profiling.
- Experience in LLM training and post-training (RLHF/DPO).
- PyTorch, Transformers, Python.
- Technical leadership and willingness to write code independently.
- Advanced English or Russian.
Will be a plus: CUDA/Triton, multimodal models, open-source contributions, scientific publications.
Conditions:
- Fully remote work, full-time employment.
- Opportunity to receive a stake in the company and grow to Head of ML.
- 28 vacation days + 7 wellness days.
- Medical support up to $1,000 per year or insurance.
- Compensation for workspace, training, and conferences.
Selection Process:
- Screening with a recruiter and 5 technical questions.
- 90-minute technical interview: live coding and ML System Design.
- Final interview with the team.
To apply: recruiter contact
