AI / ML Engineer Intern: (LLM & Agentic AI)

Role Overview

Horrazon AI is looking for an exceptional AI / ML Engineer Intern to work on Pogee, our proprietary LLM and agentic layer. We are building systems that ingest live data from social platforms to provide actionable insights. This is a high-leverage role where you will move beyond prompt engineering to work on the core mechanics of model training, fine-tuning, and agentic reasoning in a production environment.

Responsibilities

• Training & Fine-Tuning: Fine-tune open-source LLMs using LoRA/QLoRA, full fine-tuning, and preference-optimization techniques like RLHF/DPO.
• Model Fundamentals: Apply deep understanding of transformer architecture, tokenization, and training dynamics.
• Agentic Systems: Design reasoning and decision layers, including multi-step planning and tool use.
• Evaluation: Build rigorous evaluation frameworks and run controlled experiments to drive evidence-backed deployments.
• Data Engineering: Prepare high-quality datasets using a Python, Go, and PostgreSQL stack.
• Efficiency: Manage inference cost, latency, and quantization decisions.

Requirements

• Pursuing or recently completed a degree in CS, AI/ML, Data Science, or equivalent.
• Demonstrated hands-on experience training or fine-tuning LLMs using PyTorch and Hugging Face.
• Deep knowledge of transformer internals (attention, embeddings, loss functions).
• Exposure to agentic AI frameworks like LangChain or LlamaIndex.
• Strong Python skills and comfort with SQL and REST/OAuth2 APIs.
• Git fluency and ability to work in an async, remote-first environment.

Skills

• PyTorch
• Hugging Face
• Python
• PostgreSQL
• LangChain

Benefits

• Direct access to the CTO and technical mentorship.
• Production ownership from day one.
• Flexible schedule and Work from home.
• Cell phone and Internet reimbursement.
• Paid sick time.