The role
This position sits within Nebius Token Factory, our serverless platform for running and customizing open-source LLMs in production. Token Factory allows for serverless inference and fine-tuning (LoRA, full FT, RFT) backed by in-house optimizations like custom speculative decoding, quantization, cache-aware routing and dedicated endpoints. Customers come to us to move from prototype to scaled production without the cost and complexity of building and tuning their own inference stack.
We seek an experienced Senior ML Solutions Architect to support customers leveraging Nebius Token Factory's serverless inference and fine-tuning platforms for open-source LLMs across multiple modalities. In this role, you will be collaborating with clients to design and implement optimized inference workflows, build customized LLM-based solutions and architect scalable AI applications using our served models. You will also work closely with our backend team to improve our platform to match clients' needs.
You’re welcome to work from Tel Aviv, Israel.
Your responsibilities will include
- Optimize LLM inference across various modalities to drive business value and support customer goals
- Provide support in supervised and reinforcement learning fine-tuning to maximize model quality for the customers
- Design and implement LLM-based solutions using Nebius Token Factory’s inference services
- Build production-ready applications leveraging our serverless LLM APIs, including multimodal models (text, vision, audio) and domain-specific models
- Provide technical expertise in prompt engineering, RAG architectures and model selection
- Collaborate with product and engineering teams to surface customer feedback and shape the platform roadmap
- Guide customers in scaling from POC to production with a focus on performance, reliability, and cost efficiency
We expect you to have
- 5+ years of experience in ML/AI systems, with at least 2 years focused on LLMs and generative AI
- Deep knowledge of the LLM ecosystem, including model architectures and fine-tuning approaches
- Hands-on experience with:
- Running LLMs in production: deploying and operating inference workloads
- LLM fine-tuning, including supervised fine-tuning (SFT/LoRA) and data preparation/curation; experience with RL-based fine-tuning is a strong plus
- LLM evaluation: building task-specific benchmarks and offline/online eval pipelines, including LLM-as-a-judge setups
- Inference frameworks and libraries (e.g., vLLM, SGLang, TensorRT-LLM, Transformers)
- Deploying LLM-powered applications using APIs from OpenAI, Anthropic, or open-source models
- Strong Python programming skills
- Excellent communication skills, with the ability to clearly explain technical concepts to diverse audiences
It would be an added bonus if you have
- Work with multimodal AI models (e.g., vision-language, speech)
- Proficiency with DevOps tools (Docker, Kubernetes)
- Contributions to open-source ML/AI projects
Preferred technical stack
- Programming Languages: Python
- ML Frameworks and Libraries: vLLM, TensorRT-LLM, SGLang, Transformers, OpenAI/Anthropic SDKs
- MLOps and DevOps tools: Kubernetes (K8s), Docker, Git
- Cloud Platforms: AWS (SageMaker, Bedrock), GCP (Vertex AI), Azure (Azure ML)
#LI-DK1