- Design, develop, and deploy ML-powered applications that leverage large language models (LLMs) to recapitulate psychological phenomena and address complex challenges in mental health
- Optimize and steer LLM behavior using techniques like activation steering, prompt engineering, and adapter fine-tuning.
- Develop an optimized data pipeline to facilitate development, scalability, and integration with a user-facing product.
- Develop scalable and efficient ML pipelines for data preprocessing, model training, and deployment in cloud environments such as AWS.
- Integrate cloud-based tools (e.g., Bedrock, EC2, S3) to facilitate large-scale application deployment and data management.
- Monitor, evaluate, and refine ML models in production to maintain performance and reliability, leveraging tools for continuous integration and deployment (CI/CD).
- Communicate findings to a multi-disciplinary audience.
- Advanced degree (MS or PhD preferred) in Computer Science, Machine Learning, Computational Neuroscience, or a related field, or equivalent industry experience
- Proven experience deploying LLM-powered applications in production environments, including expertise in frameworks like Hugging Face Transformers, PyTorch, or TensorFlow.
- Strong knowledge of cloud-based ML tools (e.g., AWS Bedrock, Lambda, EC2) and scalable architectures for training and deploying models.
- Fluency in Python and standard ML tools and libraries (e.g. PyTorch, TensorFlow, Jax, sklearn)
- Motivated and team oriented, with an ability to thrive in a multidisciplinary environment
- Excellent communication and presentation skills
Preferred experience:
- Post-Phd/MSc work experience is an advantage but not a requirement.
- Experience building SaaS products
- Familiarity with common software development tools and best practices: Git, cloud/cluster computing, UNIX environments, shell scripting.
TBD