TLDR: Drive research on LLM solutions (modeling, JER, reasoning, agentic frameworks) and transition research into production. Mentor and guide junior researchers.
- Fast pass: Published in top-tier conferences/journals. Attach your google scholar page.
- Slow pass: Masters with strong research background, even if no publications yet.
Responsibilities
- Lead the design of research and experiment methodologies for LLM core components.
- Conduct independent research and present findings to stakeholders.
- Work with engineering teams on data collection, preprocessing, quality assurance, and augmentation to build robust training datasets.
- Apply advanced analytical techniques to derive insights from large datasets.
- Provide regular updates on scientific progress, insights, and direction to supervisors and research leads.
- Publishing in top journals and conferences is valued but not mandatory.
- Support additional program-related projects as needed.
- Collaborate with engineering teams to transition research from early-stage to applied engineering.
- Senior-level staff are expected to mentor and advise team members across projects.
Requirements
- Masters/PhD in Computer Science, Engineering, Statistics, or related fields.
- 2–4 years of R&D experience in one or more of the following areas:
- NLP with focus on LLM components (entity-relation, information retrieval, prompt engineering, reasoning, sentiment analysis, LLM vulnerabilities).
- Generative AI, especially synthetic data generation.
- Strong problem-solving and analytical skills.
- Ability to stay updated on the latest research and methodologies.
- Demonstrated drive to deliver effective solutions under tight timelines.
- Excellent communication and organizational skills.
- Ability to produce detailed technical documentation, architectural designs, and best practices.
Hybrid Role: 3 days in 2 days out.