Research Scientist

Lumicity
New York, NY

AI Research Scientist – NLP/Computational Linguistics


A well-funded applied AI startup is building advanced systems to distinguish human-created content from machine-generated content. The team combines frontier research with real-world deployment, turning cutting-edge ideas into production systems used at scale.


They’re hiring an AI Research Scientist to improve and scale core detection models. This is a hands-on role where you’ll train large neural networks, design experiments, and push model performance forward, while also shipping code into production.


What You’ll Do

  • Train and fine-tune large-scale deep learning models
  • Design experiments to improve robustness and generalization
  • Create and leverage synthetic data at scale
  • Contribute new ideas that materially improve model accuracy
  • Collaborate with engineering to deploy research into production systems
  • Share insights through technical writing or publications where appropriate


This is not a purely academic role, you’ll be expected to code, ship, and iterate.


What They’re Looking For

  • PhD in ML/AI/NLP (or equivalent research experience at a leading AI lab)
  • Publications at top-tier venues (e.g., NeurIPS, ICML, ICLR, ACL, EMNLP) or meaningful contributions to deployed AI systems
  • Strong understanding of generative models and modern LLMs
  • Experience training models in PyTorch at scale
  • Evidence your research has led to real-world impact


Bonus Points

  • Work in robustness, interpretability, synthetic data, or model evaluation
  • Background in NLP, computational linguistics, vision, audio, or other adjacent ML domains
  • Experience with distributed training or ML infrastructure


Ideal Profile

  • Early-career researcher (including strong new PhDs) who wants applied impact
  • Industry researcher who wants more ownership and product exposure
  • Someone motivated by AI integrity, trust, and responsible deployment


Small team. High ownership. Deep technical problems. Real-world impact.