WORK & VENTURES CLIENT SOLUTION / AI & SOFTWARE
Naomi for
Lincoln Tax Service.
Custom AI integration, client intake and business operations software developed by Jeff Buesing through Labor Group.
A Labor Group client engagement
From disconnected steps to a connected office.
Jeff Buesing led the design and development of Naomi through Labor Group for Lincoln Tax Service. The engagement brings together a custom public website and protected staff application, combining business-process design, frontend and backend engineering, AI integration and operational implementation.
The goal was to move beyond paper-driven, disconnected steps toward consistent intake, visible progress, safer revisions and clearer handoffs—while keeping tax judgment and consequential decisions with the office team.
Client: Lincoln Tax Service
Custom application and delivery: Labor Group
Development lead: Jeff Buesing
Platform identity: Naomi
THE DESIGNED WORKFLOW
One connected progression.
Human decisions at every consequential step.
- Inquiry
- Appointment
- Intake
- Review
- Documents
- Signature workflow
- Preparation handoff
Illustrative designed workflow, not a claim that all client records have reached each stage. “Ready for TaxWise” is a preparation handoff status—not a TaxWise API integration, return preparation, or electronic filing feature.
THE BUSINESS CHALLENGE
Keep context from getting lost between steps.
Tax-office work spans inquiries, scheduling, prior-year records, changing intake answers, document preparation and signature review. Disconnected tools and paper-driven processes make it harder for staff to see what is current, what needs review and who must act next.
Naomi provides a central workspace where connected records and action-oriented statuses make the designed office process easier to follow. Improved visibility and consistency are intended operational benefits, not measured performance claims.
DELIVERED SCOPE
Software for the actual work.
- Public website, inquiry capture and inquiry follow-through
- Appointment visibility through connected scheduling
- Customer management with linked new- and returning-client intake
- Prior-year baseline review and current-year updates
- Saved drafts, checkpoints and revision-aware editing
- Work Snapshot views for progress and next actions
- Staff review before documents and signature steps
- Generated PDF viewing, downloading and printing
- Signature review with document-status tracking
- CSV export limited to customer name and intake status
- Recoverable intake Trash with preservation checks
- Staff-facing explanations for correction or developer follow-up
ARCHITECTURE / ILLUSTRATIVE
A clear boundary between assistance and records.
Conceptual diagram only; not a screenshot or disclosure of actual internal topology. The AI information assistant supports general questions from official IRS and New Jersey sources. Sensitive customer-record workflows are separated from general conversational assistance. The office retains responsibility for tax judgment and client review.
TECHNOLOGY IN SERVICE OF OPERATIONS
Custom JavaScript
Public and staff interfaces, dynamic forms, visible progress and application behavior designed around office use.
Cloudflare Workers + D1
Serverless application services and structured data for the connected administrative workflow.
OpenAI integration
Naomi's general-information assistant, grounded in official IRS and New Jersey materials—not an autonomous tax preparer.
Google Workspace + Calendar
Connection to office documents, historical records and scheduling visibility.
SignWell
Document signature-review steps and associated status tracking under staff oversight.
OAuth + MCP experience
Earlier ChatGPT-connected application phase, subsequently evolved into the standalone staff application.
ENGINEERING APPROACH
Built for interrupted, changing, real-world work.
Rather than treating intake as a stack of unrelated forms, the application uses stateful steps and saved checkpoints. Revision checks help avoid overwriting more recent work; duplicate-operation controls and recoverable request handling help staff resume after an interruption.
Uncertain saves are surfaced explicitly, with unsaved answers preserved where possible and practical corrective guidance when a staff member needs to retry or contact development support.
QUALITY & IMPLEMENTATION
Connect the code to the office.
- Document-template mapping and verification of saved output
- Frontend/backend release coordination and compatibility checks
- Automated regression and browser workflow tests
- Synthetic test scenarios, without real client records
- Iterative refinements from day-to-day office feedback
- Human confirmation before consequential document or signature actions
Naomi’s design evolved from an OAuth- and MCP-connected ChatGPT application to a standalone administrative experience, so staff no longer need to conduct client intake inside ChatGPT.
Building software around your business?
Let's scope a connected application or responsible AI integration for the way your team actually works.