🛠 Tech Stack

💼 About This Role

You'll lead the LLM team at EverAI, architecting and shipping models that power a platform with 80M+ daily tokens. You'll optimize context windows and inference latency, driving alignment and moderation strategies for millions of users. This role offers the unique challenge of building nuanced, production-grade AI in the uncensored/NSFW space.

🎯 What You'll Do

• Write production code in Python/PyTorch and mentor the team.
• Own the core chat loop, optimizing context windows and inference latency.
• Drive SFT and RLHF/DPO strategy for model alignment.
• Design custom classifiers for nuanced content moderation.

📋 Requirements

• 8+ years of engineering experience with significant ML/LLM work.
• Shipped LLM features to millions of active users at scale.
• Proficient in Python/PyTorch and modern LLM stack (vLLM, HuggingFace).
• Comfortable working with NSFW/uncensored content and its moderation.

✨ Nice to Have

• Experience with RAG and memory retrieval systems.
• Familiarity with data sourcing and labeling for LLM training.
• Background in building safety classifiers for explicit environments.

🎁 Benefits & Perks

• 🌍 100% remote work with global impact.
• 🚀 Exponential growth – from 50M to 500M users.
• 🧠 Proven leadership with IPO and exit experience.
• 🛡️ Ethical core – build with EverGuard moderation tech.