AI Engineer, Multimodal LLMs

Sourced from feeny.ai

Eloquent AI is building multimodal autonomous AI agents for high-stakes enterprise workflows in financial services. As an AI Engineer, you will design, deploy, and optimize LLM-based agents that understand, act, and integrate with enterprise systems. You'll work on multimodal architectures combining language, speech, vision, and reinforcement learning to deliver real business impact.

• Build, deploy, and optimize AI agents for enterprise-grade conversations
• Design and develop next-gen multimodal LLM architectures (LLMs, speech, vision, reinforcement learning)
• Explore trade-offs between model quality and efficiency when translating research into practical solutions
• Refine training paradigms for real-world applications
• Integrate AI agents with enterprise systems via APIs, databases, and automation tools
• Experiment rapidly to improve AI-driven interactions, response quality, and automation capabilities
• Collaborate with cross-functional teams (engineering, research, and product) to shape the roadmap
• Monitor and improve agents' performance via user simulations and evaluations

• 3+ years of experience in software development, AI engineering, or NLP in a production environment
• Strong proficiency in Python, with experience in frameworks like PyTorch and TensorFlow
• Experience working with LLMs or large computer vision models, or generative AI models, including fine-tuning and inference optimization
• Familiarity with APIs, cloud infrastructure (AWS, GCP, or Azure), and enterprise integrations
• Ability to prototype, experiment, and iterate quickly to improve AI agents
• Strong problem-solving skills and the ability to work closely with customers to refine AI solutions
• Solid mathematical foundation of machine learning and deep learning techniques

• Experience with prompt engineering, parameter-efficient fine-tuning (PEFT), retrieval-augmented generation (RAG), reinforcement learning for LLMs
• Published AI research in top tier AI conferences like NeurIPS, ACL, SIGIR, ICML, and ICLR
• Contributed to open-source NLP projects
• Worked in a fast-paced startup environment and thrive in rapid iteration cycles