Описание
The company is building an AI agent that performs everyday tasks, manages workflows, and retains context across long conversations. Its machine learning layer is a core part of the product.
Задачи
•Build end-to-end pipelines for data, training, evaluation, and inference
•Adapt and fine-tune models using LoRA, QLoRA, SFT, DPO, and distillation
•Architect inference systems that meet latency and cost constraints
•Create pipelines for high-quality synthetic and real-world training data
•Evaluate robustness, safety, bias, and production behaviour beyond benchmarks
•Own model deployment, including GPU optimisation, quantisation, memory efficiency, and scaling
•Work with application engineers to integrate ML into backend, mobile, and desktop products
•Take ownership of ambiguous problems from zero to one
•Ship, iterate, and learn from production
•Influence how the architecture evolves
Требования
•Deep understanding of deep learning and transformer architectures
•Proven experience training, fine-tuning, or shipping large-scale models in production
•Strong knowledge of at least one major ML framework, such as PyTorch or JAX, and ability to learn others quickly
•Familiarity with distributed training and inference tools, including DeepSpeed, FSDP, Megatron, ZeRO, or Ray
•Engineering discipline in writing readable, robust, and maintainable code
•Experience optimising for GPU constraints, including quantisation, mixed precision, and memory
Nice to have:
•LLM inference frameworks such as vLLM, TensorRT-LLM, or FasterTransformer
•RLHF methods such as PPO, DPO, or ORPO
•Open-source contributions to ML or systems libraries
•Scientific computing, compiler, or GPU kernel experience
•Multimodal or diffusion model experience
•Large-scale data processing with Arrow, Spark, or Ray
Условия
No conditions specified
