Senior Machine Learning Engineer
Who we are
Nuclearn.ai is a startup technology company delivering solutions to the nuclear and utility industries. We solve meaningful problems using advanced AI and high-quality software, deployed easily and cost-effectively. Nuclearn’s latest product, AtomAssist, provides private, nuclear fine-tuned large language models to support a variety of generative AI use cases.
Nuclearn deploys its products in a web platform that tightly integrates workflow automations, machine learning model deployment, dataset management, and analysis to solve specific problems for our customers in both the nuclear and broader utility industries. The Nuclearn.ai platform is currently installed and in use at over 33 nuclear reactors in North America, and we are actively working to expand our reach to every nuclear utility in the US and abroad, as well as to other utility sectors.
We're looking for a passionate, experienced Machine Learning Engineer to take Nuclearn to the next level through the research, development, training, evaluation and deployment of state of the art machine learning models and algorithms. In this position, you will work directly with our company’s founders to develop new machine learning models that will solve our customer’s automation and analytic needs. If you thrive in a start-up atmosphere where you can have a huge influence on company direction, be surrounded by motivated people, and deliver real AI solutions using state of the art techniques, then this position is for you.
You must be a US citizen or permanent resident (e.g. green card holder) to be eligible for this position due to DOE Export Compliance requirements.
What you’ll be doing
Model Development and Fine-tuning: Fine-tune large-scale language models (>10B parameters) for new products or specific tasks. Expected techniques include:
- Deep fine tuning models like Llama and T5.
- Document Generation using deep fine-tuning, retrieval augmented generation, prompt chaining, and other techniques.
- Question & Answer over private document sets using deep fine-tuning and retrieval augmented generation.
- Training LoRA adapters and/or task specific classification heads.
- Massive (>1k labels) multi-label classification.
- Risk control and alignment using training and post-training techniques (e.g. RLHF, Conformal, etc).
- Model distillation for performance optimization.
Research and Development: Continuously research and evaluate emerging technologies and methodologies in machine learning and NLP to keep the team updated and capable of pivoting quickly.
Prototyping and Experimentation: Quickly implement proof-of-concept prototypes to validate new approaches, technologies, or algorithms.
Codebase Management: Collaborate with the engineering team to integrate machine learning algorithms into the existing codebase.
Data Management: Work closely with the team to design and maintain scalable data pipelines and architectures.
Resource Management: Optimize the computational efficiency and costs associated with machine learning tasks.
What we expect from you
- Bachelor's or Master’s degree in Computer Science, Machine Learning, or a related technical field.
- 5+ years of experience in machine learning engineering, at least 1 year of experience with large language models.
- Strong programming skills in Python and experience with PyTorch.
Bonus qualifications (not required, but a plus)
- Experience training models in multi-node clusters.
- Prior experience working in a startup environment.
- Knowledge of the nuclear or utility industries.
- Salaried position
- Mon-Fri hybrid work environment (expectation is at least 75% in office)
What you’ll love
- Competitive pay, bonus structure and equity
- Opportunity to make a huge impact in a growing, venture capital backed company
- Unique opportunity to apply state of the art machine learning techniques to new problems
- Unlimited vacation