AI Developer Document Intelligence

Springbord Systems; United States; Remote; USD 120,000 - 180,000; Full time

Job summary

Springbord Systems is building an AI-powered document intelligence capability to support our commercial real estate operations. We are looking for an AI Developer to own a production system that extracts structured data from complex legal documents, and to maintain and extend it as the platform grows.

You will take over an initial implementation from an external partner, become the in-house owner of the system, and work with our technology and operations teams to improve extraction accuracy,

Qualifications

• 3–5 years of software development experience with Python.
• Experience building applications with large language models (prompt engineering, API integration, structured output).
• Familiarity with retrieval-augmented generation concepts (chunking, embeddings, vector search).
• Experience processing documents programmatically (PDF parsing, text extraction).
• Comfortable working with REST APIs and PostgreSQL.
• Ability to read and take ownership of an existing codebase.
• Good written English and clear communication with non-technical stakeholders.

Responsibilities

• Own, maintain and extend an LLM-based document extraction pipeline in production.
• Improve extraction accuracy through prompt design, retrieval tuning and evaluation.
• Build and maintain evaluation sets and accuracy benchmarks; investigate and resolve extraction errors.
• Extend the system to cover additional document types and data categories.
• Work with PII redaction and anonymisation components to ensure sensitive data is handled correctly.
• Integrate extraction outputs with internal applications via APIs; collaborate with domain experts to translate requirements into extraction logic.
• Maintain technical documentation and contribute to code quality and testing practices.

Skills

Python

LLMs / Prompt engineering

REST APIs

Postgres

English communication

Codebase ownership

Tools

PDF parsing

Vector search / embeddings

Self-hosted models

API integration

Laravel / PHP (nice to have)

Job description

Job description

About the role

Springbord is building an AI-powered document intelligence capability to support our commercial real estate operations. We are looking for an AI Developer to take ownership of a production system that extracts structured data from complex legal documents, and to maintain and extend it as the platform grows.

You will work closely with our technology and operations teams, taking over an initial implementation delivered by an external partner and becoming the in-house owner of the system going forward.

What you’ll do

• Own, maintain and extend an LLM-based document extraction pipeline in production
• Improve extraction accuracy through prompt design, retrieval tuning and evaluation
• Build and maintain evaluation sets and accuracy benchmarks; investigate and resolve extraction errors
• Extend the system to cover additional document types and data categories
• Work with PII redaction and anonymisation components to ensure sensitive data is handled correctly
• Integrate extraction outputs with internal applications via APIs Collaborate with domain experts to translate business requirements into extraction logic
• Maintain technical documentation and contribute to code quality and testing practices

What we’re looking for

Required

• 3-5 years of software development experience, with strong Python
• Hands-on experience building applications with large language models (prompt engineering, API integration, structured output)
• Familiarity with retrieval-augmented generation (RAG) concepts chunking, embeddings, vector search
• Experience processing documents programmatically (PDF parsing, text extraction)
• Comfortable working with REST APIs and relational databases (Postgres)
• Ability to read and take ownership of an existing codebase
• Good written English and clear communication with non-technical stakeholders

Nice to have

• Experience with PII detection or data anonymisation tooling
• Exposure to evaluation frameworks for LLM outputs l Background in legal-tech, fintech, insurance or another document-heavy domain
• Familiarity with Laravel or PHP-based systems
• Experience running models locally or in self-hosted environments

Mindset

• Methodical about accuracy comfortable with the idea that "mostly right" isn't good enough on legal documents
• Curious about the business domain, not just the technology
• Able to work independently while knowing when to ask