Applied Scientist

Xenonstack Private Limited seeks an Applied Scientist in Mohali, India, to conduct research on agentic AI systems and turn findings into deployable enterprise applications. The role combines machine learning, reinforcement learning and large language models with experimentation on reasoning, multimodal intelligence and optimization. This is a full-time position with a listed salary of 8-15 LPA.

Responsibilities

• Research large language models, reasoning systems and multi-agent architectures. Prototype reinforcement learning, retrieval-augmented generation and hierarchical memory for enterprise use.
• Design and evaluate experiments to improve context management, generalization and reasoning accuracy. Produce internal research publications and model evaluation reports, and contribute to open-source initiatives.
• Work with engineering teams on deployable prototypes; build and assess agent frameworks and orchestration strategies using LangChain, LangGraph or equivalent tools. Develop and test AI safety and interpretability techniques, and integrate research advances into products.
• Collaborate with ML engineers, data scientists and AI product teams; present findings internally and stay current with research in LLMs, RLHF, multimodal AI and agentic reasoning.

Required qualifications

• A master’s or Ph.D. in Computer Science, Artificial Intelligence, Machine Learning or a related field, and 3–6 years of research experience in AI, ML or NLP. The job information section separately lists work experience as 1–3 years.
• Expertise in deep learning frameworks such as PyTorch, TensorFlow and JAX; experience with LLMs, Transformers, RAG pipelines and reinforcement learning, including RLHF and RLAIF.
• Python proficiency, familiarity with research tools such as Weights & Biases, Hugging Face and LangChain, and the ability to design, conduct and rigorously analyze experiments.

Preferred qualifications include research in multi-agent systems, symbolic reasoning or causal inference; knowledge of LLM evaluation metrics and AI safety; research publications or open-source contributions; and familiarity with distributed training and optimization at scale. The role offers collaboration with the CTO and Chief Scientist and work on enterprise-focused, responsible AI systems.