LLM system achitect

LLM Systems Architect

WhiteKnight Law Platform

The problem is simple: when fraud happens, the victim usually has evidence everywhere. Emails, contracts, screenshots, bank records, text messages, corporate filings, police reports, attorney letters, intake forms, and messy timelines.

The system is not built for them. We work with governmental crime victim advocacy groups turn disorder into a clear, organized, usable case file that a prosecutor can use.

The Role

We are looking for an LLM Systems Architect to maintain and enhance the AI layer behind case intake, evidence organization, legal advocacy services, secretary workflows, document review, and matter preparation.

This is not a chatbot role. This is a systems role.

You will help build AI infrastructure that can ingest messy information, classify it, summarize it, extract timelines, identify missing evidence, prepare intake packets, support advocates, and help legal teams understand the facts faster.

Responsibilities

• Design LLM workflows for case intake and evidence organization.
• Build RAG systems for documents, communications, and matter files.
• Create structured outputs: timelines, issue maps, evidence indexes, and intake summaries.
• Develop guardrails for sensitive legal-adjacent workflows.
• Build evaluation systems to test accuracy and reliability.
• Work with product, legal, intake, and operations teams.
• Help automate secretary and case-preparation workflows.
• Design systems that are secure, auditable, and human-reviewable.

Ideal Background

• Strong experience with LLMs, RAG, agents, embeddings, and structured extraction.
• Python and API experience.
• Familiarity with vector databases and document pipelines.
• Strong judgment around hallucination, verification, and auditability.
• Interest in legal tech, fraud response, investigations, or small business advocacy.
• Ability to build practical systems, not demos.

Why This Matters

we are the only Nonprofit that dedicated to helping small business businesses that have been defrauded.

Small businesses are often outmatched after fraud. They lack the time, money, and infrastructure to organize their case properly.

We are building a platform that helps victims get organized, tell the truth clearly, and move faster toward advocacy, legal support, and resolution.

This is AI with a serious purpose: helping defrauded businesses turn chaos into action.