SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Luminary is an AI-native platform transforming wealth transfer services for high-net-worth households. The company serves 1,000+ customers including multi-family offices, private banks, trust companies, and top law firms, helping them deliver personalized, end-to-end wealth transfer advice at scale.
You'll join a small, senior engineering team building sophisticated AI systems that extract and structure complex legal and financial documents into a household-level knowledge graph. This graph powers agent-driven workflows that help advisors navigate 60+ interconnected legal documents, trusts, businesses, and holdings.
Key responsibilities include:
- Improving accuracy on core document extraction and entity recognition, working with domain experts to identify and fix failure modes in complex, amended legal documents
- Owning the evaluation framework that validates system changes and prevents regressions—critical infrastructure for maintaining a probabilistic system
- Enhancing the shared agent harness, including tooling, parallel execution, context strategy, and orchestration for agentic advisory workflows
- Designing human-in-the-loop interfaces where suggestions are surfaced, explained, and accepted/rejected, feeding corrections back into the system
- Staying current with AI/ML developments and leveraging evals to adopt new models and capabilities
You'll work with Go, Postgres, Temporal, GraphQL, and observability tools like Honeycomb, Braintrust, and PostHog. The team ships to production multiple times daily with small, reviewable changes and invests heavily in modern tooling for leverage.
The company is led by David Barnard (former Head of Wealth Management at AllianceBernstein) and backed by 8VC, Fin Capital, Rockefeller Capital, Focus Financial Partners, and family offices. You'll have an office near Bryant Park in NYC with a strong in-office culture, but remote work is fully supported.
