AI Fund
Customer-facing Applied AI Engineer
About this role
LandingAI seeks an Applied AI Engineer to help transform unstructured documents into computable data using agentic systems and foundation models. You'll work on production-grade document understanding solutions that handle complex layouts, tables, and mixed modalities at a fast-growing, AI-native company founded by Andrew Ng.
What you'll do
- Develop and optimize agentic AI systems for document extraction and understanding
- Work on production ML infrastructure and deployment pipelines
- Collaborate on data-centric approaches to improve model quality and handling of edge cases
- Build specialized document models and integrate foundation models into real-world applications
- Evaluate and refine document understanding solutions using customer data
- Take ownership of technical problems and help shape product direction
What they're looking for
- Machine learning and AI systems design
- Visual AI and computer vision
- Foundation models and agentic systems
- Production ML infrastructure
- Python or similar programming languages
- Document processing and data extraction
- Problem-solving and technical depth
- Rapid prototyping and iteration
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
AI Fund
AI Fund builds and invests in AI-powered applications spanning healthcare, productivity, manufacturing, and developer tools. The company is hiring AI Engineers, ML specialists, and full-stack engineers to deploy production machine learning solutions, develop AI-driven platforms for chronic care and business intelligence, and contribute to open-source AI infrastructure projects.
View all jobs at AI FundLikely interview questions
- Can you walk us through a time when you had to explain a complex AI/ML concept or system limitation to a non-technical customer? How did you handle it?
- Our platform handles document understanding for complex layouts, tables, handwriting, and mixed modalities. What approaches would you take to debug when a model fails on edge cases in production?