Case Study
John Hyre, Tax Attorney & Investor
Operations rebuild for TaxReductionLawyer.com: airgapped PII redaction, intake automation, content distribution, and AI-drafted support.
“One of the things Ben will do in his fashion is you sit down and have a conversation — what does your business do, and the processes — and you walk through that, and he starts to think about what process can be automated.”
John Hyre, tax attorney and investor · from the joint webinar How I'm using AI in Business + Investing
The starting point
John runs a tax practice serving real-estate investors and high-net-worth filers. The business had four operational bottlenecks before Oaken AI engaged: client PII was being handled by tools not built for tax-grade audit requirements, the intake page on the marketing site was returning a 404 and silently losing prospects, content recorded for the audience wasn't being distributed past one channel, and routine inbound was eating attorney hours that should have been spent on returns.
None of these problems were unique. All of them compound. We took them one at a time, in priority order, with a working system in production before moving to the next.
What we built
Redactotron
Airgapped PII redaction engine for tax documents. Processes scanned and digital returns, removes SSNs, EINs, names, addresses, account numbers, and similar fields before any data leaves the practice. Schema-aware so it knows the difference between a SSN field and a refund amount field. Two-layer detection with fail-loud audit trails so misses are surfaced rather than hidden.
Why this one: Tax-practice PII leaks are regulatory exposure, not just embarrassment. Adobe redaction is not auditable. Generic AI tools leak data outside the practice. Hyre needed redaction that never touched an external API.
Status · In production. Used as the security gate before any AI-assisted work touches a client return.
Client Intake Automation
Rebuilt the practice's intake flow inside Kartra. New-client inquiries hit a working form, route to the right team member, and trigger automated confirmation. The previous flow had a 404'd intake page; prospects were churning before getting in the door.
Why this one: Every prospect who hits a broken form is a paying client lost. Fixing the intake was the lowest-effort revenue protection on the table.
Status · Live. Front door of the practice now works.
Content Distribution Pipeline
Single-source content recording flows to TikTok (primary) then redistributes automatically to YouTube Shorts, Instagram Reels, Facebook Reels, and Pinterest via Repurpose.io. Publer handles scheduling. The same architecture runs across multiple Oaken ventures.
Why this one: John records strong educational content. The bottleneck was distribution, not creation. Automating distribution made the content reach the audience without adding work.
Status · Running on a recurring cadence.
AI-Drafted Support Responses
Inbound client and prospect email is automatically classified (new inquiry, existing client question, scheduling request, spam). Routine responses are drafted using the practice's knowledge base and prepared for staff review and send. Staff reviews rather than composing from scratch.
Why this one: Every repetitive email is a tax attorney's time spent at the wrong altitude. Drafting from a knowledge base lets the human reviewer stay in the loop while removing the blank-page tax.
Status · In production for routine inbound categories.
Tax Law News Brief
Automated system that subscribes to tax law newsletters, IRS updates, court rulings, BiggerPockets forums, and industry blogs. Ingests, categorizes, and surfaces the most relevant items daily. Drafts newsletter content and social posts tied to current events.
Why this one: Staying current on tax law is a full-time job. Surfacing what matters and drafting commentary turns one hour of expert review into a published asset.
Status · Running. Feeds the content pipeline.
How the engagement runs
- Conversation first. Each system started by sitting with John and mapping the actual operation, not a hypothetical version of it.
- In-production before next system. No half-built tools. Each system is live with real client volume before the next one starts.
- Practice owns the code. Every system runs on infrastructure the practice controls. No proprietary lock-in.
- Joint thought leadership. The webinar series with John is part of the engagement. Clients see what an AI-augmented tax practice actually looks like, not a marketing version of it.
Where this engagement goes next
The five systems above are foundation. The next phase is connecting them into a single pipeline so a new client move from intake to onboarded portfolio in days, not weeks, and so the content pipeline pulls topics from real client question patterns. The architecture is in place. The connective tissue is the work.
Want a similar rebuild for your practice?
Call our AI agent for a free 48-hour assessment, or take the online intake.