Join our dynamic Research & Rapid Prototyping team as a Machine Learning Engineer, where you will play a critical role in bridging the gap between data science and engineering. Your work will be instrumental in developing and optimizing world-class AI tools tailored for the legal industry. This role requires a proactive and innovative engineer who can manage local GPU-based infrastructure, create efficient ML pipelines, and oversee inference workloads to power our AI-driven solutions.
Responsibilities
- GPU Infrastructure Management: Design, set up, and maintain robust local GPU-based infrastructure operations.
- Pipeline Development: Develop and manage end-to-end pipelines for building, training, evaluating, and deploying LLMs.
- Inference Management: Implement and manage inference workloads, ensuring scalability and reliability in production environments.
- Collaborative Development: Work closely with data scientists and software engineers to integrate machine learning models into production systems, ensuring seamless operation and high performance.
- Benchmarking and Optimization: Develop benchmarks to evaluate model performance and optimize algorithms for efficiency and accuracy in legal-specific tasks.
- Research and Prototyping: Rapidly prototype new ideas and solutions, staying at the forefront of emerging technologies and methodologies.
- Continuous Learning: Engage in continuous learning and demonstrate relevant projects through GitHub or similar platforms, showcasing your ability to learn and deliver innovative solutions.
Qualifications
- Technical Proficiency:
- Experience with GPU-based infrastructure and parallel computing.
- Very proficient in Python and core LLM libraries (e.g. PyTorch, Transformers, vLLM, WeightsAndBiases).
- Strong background in building and deploying machine learning models.
- Experience with ML infrastructure management, either via cloud providers or local infrastructure.
- Solid understanding of data pipelines, ETL processes, and model lifecycle management.
- Collaboration and Communication:
- Excellent verbal and written communication skills.
- Ability to work effectively in a multidisciplinary team environment, bridging the gap between data science and engineering.
- Problem-Solving and Innovation:
- Strong problem-solving skills with a creative and innovative mindset.
- Ability to prototype and experiment with new technologies and approaches.
- Learning and Adaptability:
- Demonstrated ability to quickly learn new technologies and adapt to changing requirements.
Nice-to-Have:
- Experience with air-gapped environments: Familiarity with working with highly sensitive data in high-security environments.
- Multi-modal Experience: Familiarity with multi-modal data and multi-model models.
- Graph databases: Familiarity with native graph databases (e.g. Neo4J) and integration with AI products.
- Custom GPU kernel development: Experience with developing custom GPU kernels for high-performance accelerated computing.
