Senior / Lead Research Scientist - Serbia
Serbia • FullTime
Posted 5mo ago
About the job
Inworld is a research lab building top-ranked real-time voice models used in large consumer-facing AI applications. We focus on research and development of state-of-the-art models, optimizing real-time inference, and creating best-in-class APIs and products. We are looking for strong individuals who learn fast, thrive in ambiguity, and can demonstrate their past work, rather than specific resume templates. We value impact and shipping results, with a bias for action and first-principles thinking.
Responsibilities
- Frame research questions and design experiments.
- Run experiments and analyze results.
- Write research papers and ship results.
- Contribute to the development of foundation models, including training, new architectures, and reward modeling.
- Develop and implement evaluation strategies for AI models, including benchmarks and LLM-as-judge approaches.
- Explore frontier AI topics such as multimodal models, agents, and tool use.
- Collaborate with US-based leadership and engineering teams.
Requirements
- Experience with foundation models (training, architectures, RL, reward modeling, scaling).
- Experience with AI model evaluation (benchmarks, eval loops, quality measurement, LLM-as-judge, failure analysis).
- Familiarity with frontier AI topics (multimodal models, agents, tool use, test-time compute, world models).
- Published research at top AI conferences (ICML, ICLR, NeurIPS, EMNLP, ACL, or AAAI).
- PhD in ML/NLP or equivalent practical experience.
- Demonstrated public work (AI side projects, interdisciplinary experiments, open-source contributions).
- Ability to take full-stack research ownership from question framing to shipping results.
- Professional fluency in English (written and spoken).