LLM Scientist - Agentic Frameworks
Job Description
Our client, an international AI development company based in New York, is currently seeking a LLM Scientist- Agentic Frameworks to lead strategic product development efforts in a fast-paced and collaborative environment.
This role will focus on advancing agent-based architectures powered by multimodal Retrieval-Augmented Generation (RAG) capabilities. The ideal candidate will have expertise in agentic frameworks and multimodal information processing using tools like CLIP, BLIP, and similar models. You will contribute to building intelligent agents that can perceive and interact with multiple types of input, such as text, image, and structured content.
Key Responsibilities
Agentic Frameworks & Architecture:
Design agentic architectures capable of autonomous task execution and coordination
Develop intelligent pipelines that support long-term memory, reasoning, and dynamic action planning
Multimodal RAG Systems:
Research and implement multimodal RAG pipelines that combine text and visual inputs
Integrate tools such as CLIP, BLIP, or equivalent models into context retrieval and generation workflows
Work on unifying structured and unstructured modalities for enhanced knowledge access
Research & Collaboration:
Work closely with engineers and scientists to build production-ready prototypes
Contribute to internal research efforts, knowledge-sharing sessions, and long-term roadmap planning
Qualifications & Skills
Strong understanding of agentic AI frameworks (e.g., LangGraph, AutoGPT) and autonomous systems
Proven experience with multimodal models like CLIP, BLIP, or equivalent
Practical expertise in Retrieval-Augmented Generation (RAG) architectures
Proficient in Python and experienced with machine learning frameworks, preferably PyTorch
Very strong English communication skills, both written and verbal (essential for global collaboration)
Comfortable communicating with both technical and non-technical stakeholders
Research-oriented mindset with the ability to convert theory into real-world systems
Self-driven, curious, and collaborative personality, comfortable in iterative and agile R&D environments
What you'll do
- Design, train, and evaluate large language models and agentic pipelines
- Build production-grade inference and orchestration systems
- Collaborate with product and engineering teams to ship AI features end-to-end
- Run experiments, benchmark models, and iterate on prompt and fine-tuning strategies
- Own model quality, latency, and cost across the stack
- Mentor engineers and contribute to internal ML platform improvements
What we're looking for
- Strong background in machine learning, NLP, or deep learning
- Hands-on experience with PyTorch, Hugging Face, or similar frameworks
- Experience deploying models to production
- Solid Python engineering skills
- Familiarity with RAG, vector databases, and agent frameworks
- Strong communication skills in English
Nice to have
- Published research or open-source contributions
- Experience with distributed training or GPU optimization