Pioneer Talent Program - Research Data Scientist

About the Role

We are looking for a Research Data Scientist to join our Algorithm team as part of Tech Seeds 2026. This is a research-first role for someone who wants their work to have real-world impact — not just papers, but systems that ship. You will push the boundaries of generative AI and large language model research, working alongside senior researchers to develop novel algorithms and model techniques that directly shape our products and platform.

What You'll Do

• Conduct original research in generative AI, large language models, and related areas such as alignment, reasoning, and multimodal understanding

• Design and run experiments to validate hypotheses, analyze results, and iterate on model architectures

• Develop and benchmark new training methodologies, fine-tuning strategies, and evaluation frameworks

• Contribute to the team's research agenda and collaborate with applied engineers to bring research findings into production

• Stay current with the latest developments in AI/ML research and identify opportunities relevant to our systems

• Contribute to internal research documentation and, where appropriate, external

What We're Looking For

• Master's or PhD in Machine Learning, Computer Science, Mathematics, or related field strongly preferred; exceptional Bachelor's candidates with demonstrable research experience will be considered

• 0–3 years of research or industry experience in ML/AI; strong research internship or lab experience equally valued

• Deep understanding of transformer architectures, language model pretraining, and fine-tuning techniques

• Proficiency in Python and PyTorch; experience running large-scale training experiments

• Strong mathematical foundations — linear algebra, probability, optimization

• Ability to read, critically evaluate, and build on academic research papers

• Intellectual curiosity and rigorous experimental thinking

Bonus Points

• Published or presented work at top ML venues (NeurIPS, ICML, ACL, ICLR, etc.)

• Experience with RLHF, DPO, or other alignment techniques

• Familiarity with distributed training frameworks and GPU cluster environments

What Makes You a Fit for Tech Seeds 2026

You think in research questions but execute like an engineer. You can move from reading a paper on Monday to running an experiment by Wednesday.