Tech Lead, LLM & Generative AI

Descrição da Vaga

Requisitos

• 8+ years of engineering experience, with significant time spent shipping ML/LLM features for millions of users
• Strong proficiency in Python, PyTorch, and modern LLM frameworks (Hugging Face, vLLM, fine-tuning pipelines, evaluation frameworks)
• Demonstrated experience with large-scale model deployment, alignment strategies, and moderation of sensitive content
• Deep understanding of model behavior, prompt engineering, temperature, sampling, and optimization for conversational AI
• Strong problem-solving skills, ownership mindset, and the ability to distinguish between shippable production solutions and academic perfection
• Excellent communication and mentorship abilities to guide a small but high-impact engineering team

Responsabilidades

• In this role, you will lead a growing team of engineers, architect cutting-edge AI solutions, and directly impact millions of users globally
• You will operate in a fast-paced, highly autonomous environment, tackling complex challenges in model alignment, moderation, and performance optimization
• This position blends hands-on engineering with strategic leadership, giving you the chance to shape both the technology and the team
• If you thrive in high-velocity settings and enjoy building production-ready AI systems that redefine user experiences, this role offers an exceptional opportunity
• Lead and mentor a team of AI engineers while actively contributing hands-on to the codebase using Python and PyTorch
• Architect, optimize, and maintain the core chat loop, managing context windows, memory, retrieval, and inference latency at large scale
• Own the full model lifecycle, including supervised fine-tuning (SFT), reinforcement learning from human feedback (RLHF/DPO), and retrieval-augmented generation pipelines
• Oversee data sourcing, labeling, and cleaning to improve model steerability, multicultural performance, and ethical alignment
• Design and implement high-precision moderation systems to manage sensitive content safely and responsibly
• Ensure production-ready solutions balance user freedom with robust safety guardrails, maintaining an exceptional user experience

Benefícios