This role is for one of Weekday’s clients Min Experience: 8+ years Location: Gurgaon JobType: full-time We are looking for a highly skilled and hands-on Founding Data Scientist to build, fine-tune, and post-train large language models (LLMs) that power autonomous AI agents. This is a founding-level opportunity to take ownership of the complete model improvement lifecycle, from training data creation and reward design to supervised fine-tuning (SFT), reinforcement learning, evaluation, and production deployment. The organisation is building an AI-powered personal assistant designed to help users complete real-world tasks, including making calls, managing bookings, coordinating services, and handling payments.
Its autonomous AI agents are designed to plan, execute, and deliver verified outcomes. You will work directly with the founding team to develop proprietary AI models and infrastructure, using real-world user interactions and task outcomes to continuously improve model performance. This is a deeply technical role suited to engineers and researchers who have hands-on experience training, fine-tuning, and deploying LLMs rather than simply building applications using third-party LLM APIs.
Requirements
Responsibilities
LLM Post-Training: Fine-tune open-weight language models such as Llama, Qwen, and Gemma using supervised fine-tuning (SFT), Direct Preference Optimization (DPO), and reinforcement learning techniques. Reinforcement Learning: Design reward functions and optim