Machine Learning Resident - Client: ModeDi (12 month term)
Core
Build a secure LLM-based system combining internal and public data to support real estate appraisal workflows, including highest and best use analysis and document understanding.
Role type
Applied Machine Learning Researcher (LLM Systems)
Builds
A production-ready LLM application for real estate appraisal professionals
Domain
Real Estate / Appraisal / Large Language Models
Deliverable
production ML models
Required skills
Large Language Models, Retrieval-Augmented Generation, Document Parsing, Python, PyTorch, Hugging Face, vLLM, LangGraph, Uncertainty Quantification
Preferred skills
LLM Evaluation Frameworks, Agent Orchestration, Prompt Injection Guardrails, Real Estate Domain Knowledge, Production Deployment
Technologies
PyTorch, Hugging Face, Transformers, vLLM, Unsloth, LangGraph, Pydantic, Anthropic, OpenAI
Responsibilities
Design and build secure LLM systems for appraisal workflows; Develop ingestion pipelines for municipal documents; Build citation-grounded generation and confidence estimation; Implement secure infrastructure with role-based access control; Conduct applied research in RAG and document understanding; Collaborate with stakeholders to develop MVP solutions.
Seniority
Researcher (MSc/PhD level)