PhD Research Intern

About the internship: Selected intern's day-to-day responsibilities include: 1. Driving core research in specialized areas, including LLM post-training (RLHF, GRPO, instruction tuning) and data efficiency framework design. 2. Formulating and constructing domain-specific evaluation benchmarks to test LLM capabilities across security, reasoning, and safety thresholds. 3. Designing and running rigorous technical machine learning experiments to validate research hypotheses and iteratively optimizing model concepts. 4. Collaborating hand-in-hand with research scientists and cross-functional engineering teams to rapidly prototype and evaluating novel AI approaches. 5. Authoring and producing publication-ready scientific papers targeting elite, top-tier AI/ML conferences. 6. Advancing ongoing research tracks like Agentic Quality Evaluation, Few-Shot Grounding, and Human-in-the-Loop optimization to minimize manual data annotation check barriers. Who can apply: Only those candidates can apply who: are available for full time (in-office) internship are available for duration of 3 months have relevant skills and interests Stipend: INR₹ 1,20,000 - 1,60,000 /month Deadline: 2026-09-08 23:59:59 Skills required: Python, Machine Learning, Problem Solving and Research and Analytics Other Requirements: Only those candidates can apply who: 1. Are currently pursuing a Ph.D. in Computer Science, Machine Learning, Natural Language Processing, or a highly related quantitative technical field. 2. Possess at least one prior publication footprint in elite conferences like NeurIPS, ICML, ICLR, ACL, or equivalent setups. 3. Are available to work fully on-site at Uber's corporate facility in Hyderabad, India. About Company: Uber's mission is to reimagine the way the world moves for the better. Uber AI Solutions acts as an internal engine designed to build extensive AI-driven platforms and robust data structures that impact millions of global users daily. Additional information: 1. Deep expertise in LLM post-training workflows (RLHF, instruction tuning), AI agents, alignment, or safety systems. 2. Exceptional coding proficiency in Python paired with extensive hands-on experience using deep learning frameworks like PyTorch or JAX. 3. Proven history of fine-tuning or training complex large language models from scratch or via open-source variants. 4. Strong scientific publication record at tier-1 machine learning avenues (e.g., NeurIPS, ICML, ICLR, ACL, EMNLP, CVPR, COLM). 5. Familiarity with distributed training infrastructure frameworks (e.g., DeepSpeed, FSDP, Megatron) is highly preferred. 6. Strong structural problem-solving skills to bridge theoretical generative AI concepts with real-world business data pipelines.