AI Engineer, Loop

• 1+ years of hands-on experience in deep learning frameworks (e.g., PyTorch, Tensorflow, etc.)
• Solid understanding of fundamental ML algorithms, especially multimodal LLMs
• Experience fine-tuning LLMs and deploying them to production or building out agentic systems
• Ability to ship high-quality code to production
• Experience in cloud environments (AWS, Google, Azure)
• Strong communication skills and willingness to collaborate in a cross-functional team environment with domain experts
• Keep up to date with the latest AI research
• Ability to work in a fast-paced environment

Desirable

• Loop is growing its AI team and you’ll have the opportunity to build both AI models and features that directly impact Loop’s business
• Loop is positioned to disrupt the incumbent freight audit and pay market, and is one of the companies leading the AI services wave, utilizing AI to automate complex back-office workflows and tasks
• You will face and solve many complex technical challenges while you receive guidance and feedback from the team
• The range of work here is broad, you can work on everything from training and deploying in-house multimodal LLMs, scaling our inference infrastructure, or building out and shipping AI agent workflows
• In doing so, you’ll have the opportunity to define how the AI and broader Loop team will grow
• Our primary focus has been on document extraction and understanding, where we utilize multimodal LLMs to extract, normalize, and link data together into our domain model
• As Loop’s customers rely on Loop to ingest and normalize highly accurate data, we hold ourselves to a high standard to build models with a very high level of accuracy
• In tandem, Loop’s machine learning platform requires a high degree of reliability and scalability, and we expect our training and inference volume to scale several orders of magnitude in the coming year
• Going forward, Loop will expand its AI capabilities, expanding into other areas such as workflow automation and audit, where we will utilize agents to tackle these problems
• ML modeling – training, evaluating, and deploying models
• ML infrastructure – scaling data infrastructure, training, or inference, improving reliability of ML systems at Loop
• AI engineering – utilizing and orchestrating API LLM models to solve business problems at Loop
• Backend engineering – building out atomic tasks, general backend work in the servicing or automation domain
• Scaling up throughput of Loop’s inference engine through continuous batching
• Developing ways to fine-tuning multimodal LLMs to reduce hallucinations
• Build out agents that audits freight invoices