We build transparent AI for California Workers' Compensation claims professionals. Our software affects real people's financial outcomes. Correctness over speed. Explainability over magic. Human in the loop, always.
Our Mission
Every AI product in the legal and insurance space faces the same temptation: optimize for impressive outputs, hide the reasoning, and hope the user trusts the result. We built Glass Box Solutions to reject that approach entirely.
AdjudiCLAIMS is built on a single design principle: every output must be traceable, citable, and understandable. The AI shows you what it found, where it found it, and what the statute says about it. Not because we have to — because that's the only kind of AI worth building for professionals whose work affects real people's financial outcomes.
We started in California workers' compensation because it's one of the most complex, most under-served, and most consequential domains in the country. Claims examiners make hundreds of decisions per week that affect injured workers' medical care and financial security. They deserve better tools.
Correctness over speed
We do not ship features that aren't fully tested. 100% passing tests is a requirement, not a goal. Claims data affects real people's financial outcomes.
Explainability over magic
Every AI conclusion must show visible, auditable reasoning. Confidence scores alone are not explanations. Every calculation shows its formula.
Human in the loop, always
AdjudiCLAIMS assists — it never decides. The examiner makes every substantive judgment call. The AI is the advisor, not the adjudicator.
Compliance by design
UPL boundaries are enforced by code, not policy. Security is architecture, not a checklist. The legal framework is built into the product, not bolted on.
Simplicity over cleverness
The most useful features are the ones that work reliably, every time. Complexity is the enemy of adoption by an examiner managing 150 open claims.
Team
Alex Brewsaugh
Founder & CEO
Alex brings deep experience in California workers' compensation law and claims operations. He founded Glass Box Solutions to solve the training and compliance problems he saw firsthand in the industry — and to prove that AI can be both powerful and fully transparent.
Sarah Brewsaugh
Chief Legal Officer
Sarah is a licensed California attorney with extensive workers' compensation experience. She leads UPL compliance strategy, reviews every AI prompt and zone boundary, and serves as the legal authority on what AdjudiCLAIMS can and cannot say. Architect of the 46-document legal framework.
Matt Ward
Chief Technology Officer
Matt leads engineering at Glass Box. His background spans enterprise software, distributed systems, and applied ML. He designed the three-stage UPL enforcement pipeline and the Glass Box explainability architecture.
Brian Barb
Chief Operating Officer
Brian has spent his career building software for regulated industries. At Glass Box he translates the complex world of claims compliance into clear, usable workflows — ensuring the product is the training program, not just a tool.
California workers' compensation is one of the largest, most complex insurance systems in the world. It covers 18 million workers. It generates over $22 billion in annual premiums. It operates under a regulatory framework of extraordinary complexity — dozens of statutory deadlines, hundreds of regulatory requirements, and a litigation environment that punishes every missed step.
Yet the software available to claims examiners is largely unchanged from a decade ago. Most systems track status but don't assist with decisions, don't surface the regulatory context that explains why each step matters, and don't teach examiners anything along the way. We're fixing that.
$22.5B
CA WC Annual Premium
127%
Combined Ratio (2024)
$4.2B
Total ALAE Annually
25-40%
Annual Examiner Turnover
Legal Framework
Our CLO built the most comprehensive UPL compliance framework of any early-stage legal AI company. Every AI boundary is documented, reviewed by counsel, and enforced by code.
Cal. Bus. & Prof. Code § 6125
Unauthorized Practice of Law — the foundational constraint governing every AI output in AdjudiCLAIMS.
HIPAA Technical Safeguards
PHI handling, audit logging, access controls, and breach notification designed for covered entity compliance.
CCPA / CPRA
Right to access, right to deletion, data minimization, and opt-out rights implemented in the product.
California Ins. Code
Unfair claims practices (Ins. Code §§ 790-790.10), fair dealing standards, and audit obligations built into every workflow.
CA Labor Code
LC 4650 TD payment deadlines, LC 5402 compensability presumptions, LC 3762 retention requirements — all enforced in the product.
SOC 2 Type II
Controls across CC6 (Logical Access), CC7 (System Operations), and CC8 (Change Management) implemented from day one.