Tech Lead (LLM / AI)

About the Role

We are building a large-scale conversational AI platform focused on personalized human–AI interaction. The product operates at significant real-world scale, serving tens of millions of users and processing over 80 million tokens per day. The platform works in a complex domain that requires advanced alignment, moderation, and low-latency inference.

We are looking for a Tech Lead to take ownership of the LLM direction and lead a small team of AI engineers. This is a hands-on role combining system design, model development, and production engineering. The product operates in an uncensored / NSFW environment, which requires careful, context-aware moderation rather than simplistic safety filters.
 

What You Will Do

• Lead and build the LLM systems powering the core conversational experience. You will architect and evolve the core chat loop, including context windows, memory and RAG pipelines, and inference performance.
• Own the full model lifecycle, including prompt design, supervised fine-tuning, and preference optimization. You will make decisions on when to fine-tune models, when to redesign retrieval pipelines, and how to manage data quality at scale.
• Design and implement moderation and alignment systems suitable for an explicit environment. This includes building custom classifiers and moving beyond binary safe/unsafe decisions toward nuanced, context-aware moderation.
• Work closely with DevOps and backend teams to deploy, monitor, and continuously improve production systems used by millions of users globally.
   

Requirements

• You have 8+ years of engineering experience and a strong track record of shipping ML or LLM-based systems to production at scale.
• You are highly proficient in Python and PyTorch and familiar with modern LLM tooling and infrastructure.
• You are comfortable working with NSFW content and understand the technical and ethical challenges of moderating explicit conversational systems.
• You have strong intuition for LLM behavior, including how prompting, sampling, and alignment techniques affect model outputs.
• You prioritize practical, shippable solutions and take ownership of system reliability, performance, and user impact.
   

Team & Environment

You will lead a small LLM team and collaborate closely with DevOps and web/backend engineers. The environment emphasizes autonomy, direct communication, and a strong bias toward execution and measurable impact.