Hamachi.ai has been awarded U.S. Patent No. 12,641,064 for an agentic AI system that redacts personally identifiable information before it reaches a large language model, the company announced Tuesday. The patent covers secure, multi-agent communications designed specifically for wealth management firms that handle account statements, tax forms, trust documents, and other sensitive client data.
The Encinitas, California-based company built its platform around a single constraint: client PII should not be the price of using AI. Its architecture identifies and strips sensitive financial data before any model processes it, preserving enough context to generate useful drafts while keeping raw client information out of downstream LLMs entirely.
How the architecture works
The patented system ingests data from email, CRM platforms, and other enterprise systems while preserving workflow context. It then orchestrates specialized AI agents based on the communication type, generates LLM-assisted drafts, and reviews outputs for errors, omissions, and compliance issues. Secure restoration, compliance gating, and audit logging happen inside the advisor's existing environment.
"AI is going to transform wealth management, but the industry needs a safer default," said Mike Wilson, CEO and co-founder of Hamachi. "Account statements, tax forms, trust documents, CRM records and portfolio data should not be casually uploaded into general-purpose LLMs. Sensitive client data should be protected before the model ever sees it."
Industry response to privacy-first design
For regulated firms, the tension between AI productivity and compliance risk is real. Aaron Spradlin, chief technology and security officer at United Planners, said Hamachi's approach "is not asking advisors to choose between innovation and control. It is bringing AI into advisor workflows with the privacy and compliance architecture firms like ours need."
Hamachi positions this patent as the first in a broader portfolio covering privacy-first, compliance-aware agentic AI systems. Additional patent applications are pending, targeting both wealth management and other regulated industries. The company was founded in 2025 by fintech veterans from Orion, Redtail, AdvisoryWorld, and Advizr.
For communications professionals watching the intersection of AI for PR & Communications and regulated industries, Hamachi's architecture signals a shift toward infrastructure that treats compliance not as an afterthought but as a pre-processing requirement. The same principle applies across AI for Finance, where client data protection and regulatory guardrails determine whether a tool gets adopted or blocked.
Why this matters for PR and communications professionals
Financial communications teams face a specific bind: they need AI speed, but they cannot risk exposing client data or running afoul of SEC and FINRA rules. Hamachi's patent describes a practical workaround - redact first, generate second, restore securely - that removes the most common objection to using LLMs on sensitive material. For PR leads inside banks, RIAs, and asset managers, this kind of architecture makes it possible to adopt AI drafting and review tools without triggering compliance alarms or requiring clients to opt into data exposure they never agreed to.
Your membership also unlocks:
