Teams at Infrrd

Build and debug Python features and utilities for an intelligent document processing product, implement GenAI/LLM workflows (LangChain/OpenAI), design PGVector-based retrieval, create evaluators and test harnesses, instrument logs/metrics, triage and ship fixes, harden deployments, and maintain SOPs and runbooks for integrations and compliance-sensitive flows.

Manages a team of residential mortgage underwriters and QC analysts, overseeing loan underwriting, post-closing quality control, compliance, escalations, staffing, training, and performance. Reviews complex VA, FHA, and Conventional loans; ensures adherence to agency, investor, and regulatory guidelines; leads root-cause analysis and CAPA initiatives; tracks quality, turnaround, productivity, and defect metrics; and supports internal, investor, HUD, and VA audits.

Build and own high-throughput Java backend services for mortgage document processing. Design asynchronous pipelines, LLM integrations, resilience and failover logic, and production observability. Optimize accuracy, latency, and cost while debugging production issues. Review code, write design documents, decompose features, and mentor junior engineers. The role requires strong concurrent Java architecture, Spring Boot microservices, NoSQL and relational databases, containerization, Kubernetes, and cloud experience.