Project Manager, Applied AI

LILT is an AI company focused on making information accessible through machine translation, human-in-the-loop expertise, and language technology. The Project Manager, Applied AI will lead large-scale multilingual data collection and LLM evaluation initiatives, coordinating global annotator and data specialist teams. The role focuses on project delivery, quality assurance, KPI monitoring, stakeholder coordination, and workflow optimization.

Responsibilities:

• End-to-End Delivery: Manage the full lifecycle of AI data projects, from scoping and guidelines creation to data delivery and post-mortem analysis
• Pipeline Management: Oversee large-scale data pipelines for multilingual data collection (audio, text, image) and LLM evaluation (RLHF, SFT, ranking, and safety testing)
• KPI Tracking: rigorously monitor and report on key performance indicators, including:
• Throughput: Volume of data processed per hour/day
• Quality: Accuracy scores, Inter-Annotator Agreement (IAA), and gold-set performance
• Productivity: Cost-per-task and worker efficiency rates
• Quality Control: Run QA loops, root-cause analysis for quality dips, and corrective training for annotator pools
• Dashboards: Maintain dashboards to visualize project health and flag bottlenecks in real-time
• Global Coordination: Manage relationships with data experts and crowd pools, ensuring adherence to SLAs regarding localized nuances and linguistic accuracy
• Cross-Functional Collaboration: Liaise with Applied AI Technical Ops teams. Translate technical requirements into clear, actionable guidelines for non-technical annotators
• Feedback Loops: Facilitate continuous feedback loops where data insights drive updates to annotation guidelines and model fine-tuning strategies

Requirements:

• • Experience: 3-5+ years of project management experience, specifically within AI/ML data operations
• • LLM Knowledge: Strong understanding of LLM training processes (Pre-training, SFT, RLHF) and evaluation methodologies (Human-in-the-loop, red teaming)
• • Data Proficiency: Advanced proficiency in Excel/Google Sheets; ability to write SQL queries to extract and analyze performance data
• • Methodology: Proven track record using Agile, Scrum, or Kanban methodologies to manage complex workflows
• • Communication: Exceptional ability to write clear, unambiguous guidelines for multilingual audiences
• • Multilingual: Fluency in a second language is highly desirable
• • Technical Tools: Experience with data annotation platforms (e.g. Scale AI, Super Annotate) and project management tools (e.g. Jira)
• • Education: Background in ML Engineering, Computer Science, Data Science and Project Management training