Overview Be among the first 25 applicants At Sonder, we believe that every person deserves to feel safe, supported, and empowered to be at their best — wherever they are. Sonder’s mobile platform provides 24/7, real-time support from a dedicated team of safety, medical, and mental health professionals, ensuring immediate care when needed. In critical situations, we offer on-the-ground assistance to manage time-sensitive incidents. By harnessing real-time insights, we enable organisations to take proactive steps toward a healthier, safer future with tangible bottom-line outcomes. Role Summary Are you a recent PhD graduate or Postdoctoral researcher passionate about applying cutting-edge AI to real-world problems? Join Sonder as an Applied Machine Learning Scientist/Engineer — focusing on LLM Research & Fine-Tuning Engineering, to build a new generation of AI systems that directly impact human well-being and safety. This role is distinct from typical AI tooling or RAG-based roles. We are not seeking candidates whose primary experience is in building vector search systems or integrating off-the-shelf APIs. We require deep, hands-on expertise with the internal workings of large language models. Your work will be instrumental in developing our personalisation engine and fine-tuning state-of-the-art open-source LLMs like LLaMa, Mistral, and DeepSeek. Your research will translate into specialised LLMs that provide empathetic, life-saving support to our members. Key Responsibilities Research, fine-tune, and optimise open-source LLMs using techniques such as PEFT, LoRA, and QLoRA to build proprietary AI models. Design and implement personalisation and recommendation engines to deliver tailored, proactive support. Develop and maintain robust machine learning pipelines for model training, evaluation, and deployment. Collaborate with cross-functional teams (product managers, software engineers, mental health professionals) to translate research into product features. Stay current with AI and LLM advancements; apply new techniques to improve systems. Publish and/or present research findings at conferences or in journals where appropriate. Key Qualifications Recently completed PhD or Postdoctoral experience in Machine Learning, AI, Computer Science, or related field. Deep understanding of Transformer architectures, attention mechanisms, and training dynamics. Hands-on experience fine-tuning open-source LLMs (e.g., LLaMa, Mistral, DeepSeek). Strong CS fundamentals and Python proficiency. Experience with deep learning frameworks such as PyTorch or TensorFlow. Passion for applying AI to high-impact domains such as mental health and well-being. Candidates with primary experience limited to RAG systems or MLOps tooling are not a fit for this role. Nice-to-Have Experience with distributed training for large models. Familiarity with cloud-based ML stacks (e.g., AWS, GCP). Track record of peer-reviewed publications in AI/ML conferences. Impact & Growth Your work will directly contribute to saving lives and improving mental well-being on a global scale. You will shape our core AI technology from the ground up and grow as a leader in a mission-driven environment. You’ll have freedom to explore new ideas and resources to deploy them, making a tangible difference. Why Join Sonder? Hybrid and flexible working environment for balance of in-office collaboration and remote work. Free access to the Sonder app for you and your family for 24/7 support. Development budget and study leave; parental leave with return-to-work policy. Employee Share Option Plan (ESOP) and salary packaging options. Wellbeing stipend and annual allocations for wellbeing activities. Coffee, catered lunches, team social funds, and dog-friendly office. Volunteer leave: 2 days per year for community involvement. We are committed to a diverse and inclusive workplace. If you need support with your application, please reach out. Due to the nature of our industry, all employees are required to complete a National Police Check and Working With Children Check. Legal and Diversity Note Sonder is for Everyone. We encourage applications from people of all backgrounds, experiences, abilities, and identities. J-18808-Ljbffr
