Machine Learning Lead (LLM)
B
Blue Rose ResearchPolitical Tech
Remote, primarily East Coast hoursFull-TimeLead
Salary165,000 - 210,000 USD per year
Apply NowOpens the employer's application page
Job Details
- Experience
- 1+ years leading data science teams; 6+ years in ML or data engineering
- Required Skills
- DockerPythonSQLKubernetesPyTorchTensorflowTerraform
Requirements
- 1+ years leading data science teams; 6+ years in ML or data engineering.
- Strong background in applied statistics, model selection, tuning, and evaluation.
- Proficient in Python, SQL, and modern ML frameworks (PyTorch, TensorFlow, or JAX).
- Experienced in building and deploying production ML and deep learning pipelines.
- Familiar with LLMs, embeddings, agentic workflows, and RAG systems.
- Comfortable with cloud and DevOps tools (Docker, Kubernetes, Terraform).
- Skilled in exploratory data analysis and handling imperfect real-world data.
- Collaborative leader who communicates clearly with technical and nontechnical teams.
- Mission-driven, curious about civic and political applications of AI.
Responsibilities
- Lead a team of senior data scientists focused on fine-tuning large language models, conducting cutting-edge R&D, and building production inference systems.
- Collaborate with senior leadership to define the team roadmap and align priorities with organizational goals.
- Lead weekly meetings and standups, keeping the team unblocked and execution moving forward.
- Provide technical direction across projects using open-weight and off-the-shelf LLMs, as well as other advanced ML techniques.
- Oversee experimentation, optimization, and data quality to ensure models are accurate, reliable, and production-ready.
- Foster creative problem-solving and methodological rigor when challenges require custom solutions beyond standard ML approaches.
- Translate complex model outputs into actionable insights for stakeholders, ensuring technical work drives real-world impact.
View Full Description & ApplyYou'll be redirected to the employer's site
