AI/ML Engineer (Backend) Engineer

ABOUT US
LawNet Technology Services (LTS) is the technology company behind LawNet, Singapore’s leading portal for legal research, information and transactions. An indispensable tool for the legal community since 1990, LawNet is subscribed by a majority of Singapore lawyers and is also accessible by anyone outside the profession. Users can conduct research on Singapore primary legal materials (Singapore Law Reports, unreported judgments and legislation) and secondary materials (such as Parliamentary reports, legal news, textbooks and journals). LawNet continues to enhance its services and content while maintaining its affordable and highly competitive subscription rates, making it an essential resource for the legal community.

LTS is a wholly owned subsidiary of the Singapore Academy of Law (SAL), a promotion and development agency for Singapore’s legal industry. In addition to running LawNet, LTS manages the technology driving SAL’s support services for Singapore’s legal industry and statutory functions such as stakeholding services and appointment of Senior Counsel, Commissioners for Oaths and Notaries Public.

Led by a Board of Directors who understands both the capabilities of technology and the needs of the legal profession, LTS continues to develop bold and innovative products and services that will better serve the needs of the legal community.

JOB DESCRIPTION

POSITIONAI/ML (Backend) Engineer

REPORTING STRUCTURE 
The AI/ML (Backend) Engineer will report to the Senior AI Solutions Architect of LawNet Technology Services (LTS).

ABOUT THE ROLE
As an AI/ML (Backend) Engineer at LTS, you will play a crucial role in developing and enhancing our AI-driven products and services. We're seeking a talented Software Engineer to join our team and be responsible for deploying, maintaining, and scaling the infrastructure that serves our custom LLMs for a variety of use cases.

RESPONSIBILITIES

• Design, develop, and implement highly scalable backend services on AWS to support efficient training, deployment, and inference of Large Language Models (LLMs) and Generative AI applications.
• Manage and optimize model runtime environments, including Amazon Bedrock, AWS SageMaker, and other model serving platforms, ensuring high availability, scalability, security, and cost efficiency.
• Keep abreast of developments in AWS and emerging AI platforms, identifying opportunities to improve model training, inference, orchestration, and operational efficiency.
• Develop and optimize APIs and service integrations to facilitate communication between AI models and frontend or downstream enterprise applications across a variety of use cases.
• Design and implement data pipelines for efficiently moving, processing, and managing training, fine-tuning, evaluation, and inference datasets using AWS services such as S3, Sagemaker, and Glue.
• Integrate containerization technologies such as Docker and container orchestration platforms to support scalable and reliable AI model deployment and inference environments.
• Define, implement, and maintain AI model evaluation frameworks, including performance metrics, benchmarks, and success criteria for assessing model quality, accuracy, relevance, latency, and cost.
• Prepare, curate, and manage evaluation datasets, including synthetic datasets, ground-truth test sets, and benchmark corpora to support continuous model validation and improvement.
• Analyse model evaluation results and develop comprehensive reports and dashboards that communicate model performance, risks, trends, and recommendations to technical and business stakeholders.
• Configure and implement robust security practices within backend services and AI platforms to protect models, data, and user information both at rest and in transit.
• Develop and implement logging, monitoring, and observability solutions to track the health, performance, and usage of backend infrastructure and AI services.
• Collaborate with AI Engineers, Data Scientists, and Frontend Engineers to understand business and technical requirements and translate them into scalable AI-powered solutions.
• Collaborate with MLOps Engineers to automate the deployment, testing, monitoring, and lifecycle management of AI models and supporting backend services.
• Stay current with advancements in Generative AI, foundation models, prompt engineering, model evaluation methodologies, and AWS AI services, including Amazon Bedrock.

SKILLS & QUALIFICATIONS

• Bachelor’s or Master’s degree in Computer Science, Artificial Intelligence, Linguistics, or a related field.
• Proven experience in NLP or related areas, with a strong portfolio of projects demonstrating expertise in data engineering and model training/finetuning.
• Solid understanding of machine learning algorithms, language modeling, and their applications.
• Proficiency in programming languages such as Python, and experience with NLP libraries (e.g., NLTK, spaCy, Hugging Face).
• Familiarity with popular vector database platforms like OpenSearch, Neo4j, or Milvus would be beneficial.
• Experience with cloud computing platforms and services, and the ability to deploy and manage large models in a cloud environment.
• Experience with data pipelines and data processing tools on AWS (e.g. AWS Lambda, Sagemaker).
• Experience working with foundation model platforms such as Amazon Bedrock, Azure AI Foundry, Google Vertex AI, or similar managed AI services.
• Knowledge of LLM evaluation techniques, including automated evaluation, human-in-the-loop evaluation, synthetic data generation, benchmarking, and A/B testing.
• Experience defining and tracking AI quality metrics such as accuracy, hallucination rate, grounding, relevance, toxicity, latency, and cost-performance trade-offs.
• Strong problem-solving skills, with a creative and analytical approach to tackling complex challenges.
• Excellent communication and collaboration skills, with the ability to work effectively in a team environment.

The successful applicant will be offered a two-year contract. Kindly note that only shortlisted candidates will be contacted.