Reducing Friction in Professional Licensing Caseloads

Executive summary

Professional licensing agencies have a dual obligation: protect the public from unsafe practice and administer complaints, licensure, and discipline through timely, fair, transparent procedures. Public documents show real operational strain. As of January 31, 2025, the Arizona State Board of Nursing had 2,177 open complaints, including 1,543 open more than 180 days; California's RN board reported 5,330 public complaints in FY 2024–25 and 2,060 complaints pending in its complaint-intake table; Maryland reported 1,296 open complaints, including cold cases, at the end of FY 2025; and Oregon's psychology board reported that only 18% of complaints were presented to the board within 180 days in its 2025 reporting year, versus a 75% target. These are different indicators with different reporting periods and cannot be added into a national backlog estimate. [1], [2], [4], [6]

The case for BoardWise is specific rather than sweeping. Its existing consumer-facing product describes guided notice interpretation, jurisdiction-aware educational resources, document organization, timelines, checklists, exportable response packets, and optional contact with independently contracted attorneys, whose services may carry separate costs. These could plausibly decrease avoidable clarification cycles and help licensees supply legible, relevant, organized information. BoardWise has not publicly demonstrated, through an independent controlled evaluation, that these features reduce agency workload; that is the hypothesis a pilot should test. [17], [18]

A regulator partnership should begin with published-process education and completeness support, not automated credibility assessments, case prioritization, legal conclusions, adverse actions, or privileged agency access. Public agencies must remain independent. A rigorous pilot would measure administrative rework, staff handling time, first-pass completeness, accessibility, factual accuracy, escalation and fairness indicators; it would expand only if benefits are demonstrated without weakening public protection or procedural rights. [12], [13], [18]

1. Research approach and evidentiary limits

This report reviews regulator annual reports, oversight audits, budget requests, regulatory surveys, and public-sector technology and accessibility guidance available as of November 1, 2026. Priority is given to primary state and federal sources. BoardWise product information is labeled as vendor-described rather than independently audited. Sources appear as clickable numbered links in the text and in the reference register below.

Important distinctions: "complaints received," "public complaint intake," "pending complaints," "unassigned investigations," "open cases," "disciplinary actions," and "cases older than 180 days" are not interchangeable. A board's licensure count is not automatically a count of unique people, and case totals across agencies cannot produce a reliable national rate without standardizing definitions, timeframes, and denominators. The findings also do not imply wrongdoing in individual matters or prove that administrative delays are attributable to any one cause. [1], [2], [3], [4], [5], [6], [7], [8]

This is a policy and product-fit analysis, not a procurement finding, third-party security certification, legal opinion, or evaluation of BoardWise in a live agency setting.

2. The caseload evidence: five concrete signals

2.1 Arizona: backlogs are simultaneously a safety and fairness problem

The Arizona Auditor General reported that, as of January 31, 2025, the Board of Nursing had 2,177 open complaints; 1,543 (71%) had been open more than 180 days, including 51 high-risk complaints, and the open complaints had been open an average of 639 days. The audit warned that delay can leave potential safety risks unresolved, erode witness recollection, defer public disciplinary visibility, and extend uncertainty for licensees whose allegations ultimately are not substantiated. It also found shortcomings in comprehensive milestone tracking and time frames. These findings support two complementary goals: prioritize genuinely dangerous cases and reduce non-investigative friction elsewhere. [1]

2.2 California: intake growth and complexity

California's RN board reported 5,330 public complaints in FY 2024–25, compared with 3,682 in FY 2021–22; the "Complaints Pending" row of its complaint-intake table increased from 1,324 to 2,060 across the same fiscal years. The FY 2024–25 "Total Received" figure of 8,317 additionally includes 1,360 convictions/arrest and 1,627 applicant items and must not be mislabeled "public complaints." A January 10, 2026 Governor's Budget proposal separately reported a 193% increase in pending unassigned investigations from 2020–21 to 2024–25, a 64% increase in complaints referred to the Investigations Division since 2020–21, and workload complexity arising in part from medspa and IV hydration matters, which the proposal notes are not yet separately tracked. The board requested eight Special Investigators; the proposal expressly did not include an IT component. The cited proposal documents the request, not its eventual enactment. [2], [3]

The sensible inference is not that an educational tool can investigate multi-subject clinical events. Rather, a regulator could explore whether better respondent-facing instructions and usable submission packets remove low-value steps from a strained process while investigators concentrate on evidence and public-safety judgment.

2.3 Maryland: administrative redesign can matter

Maryland's FY 2025 nursing report documented 930 new complaints and 1,296 open complaints including older "cold cases" (cases received in fiscal year 2019 or earlier). A January 27, 2026 Department of Health legislative briefing reported a decrease in total cases from 3,713 on January 1, 2024 to 1,167 on January 21, 2026, a count that includes administrative closures of older cases. The briefing described staffing, triage delegation, and complaint review reforms, with some tracking improvements still being implemented; it stated intake-to-triage time fell from 58 days in an earlier evaluation to under 24 business hours. These are state-reported process gains, not evidence of BoardWise effectiveness, and the numbers reflect particular snapshots and definitions: the annual report counts open complaints while the briefing counts cases, so the two are not a reconciled series. [4], [5]

2.4 Oregon psychology: cross-professional demand

Oregon's Board of Psychology reported that 18% of complaints were presented to the board within 180 days in its 2025 reporting year (reflecting investigations opened in 2024), down from 25% in 2024 and below a 75% target. The board identified case complexity, witness and attorney scheduling, waiting for records, emergency cases, respondent response and record-production time, and, as the most significant factor, unexpected investigative staff absences. Its report explains why an intervention must be tested at the correct bottleneck: a perfect submission checklist will not fix expert-review queues or staffing absences. [6]

2.5 Jurisdictional fragmentation is a design requirement

The NCSBN 2025 Discipline Survey includes 59 nursing-board jurisdictions; 37 report preponderance of the evidence as the standard of proof for final adjudication, 17 report clear and convincing evidence, and the remainder report other answers. A product cannot safely imply uniform rules for notice response, hearings, reporting or discipline. Relevant primary statutes, regulations, and current agency notices must control over any generalized summary. [7], [8]

3. Where avoidable workload enters the process

A typical agency journey is not one task but a sequence: complaint or audit receipt; intake validation; screening or priority assignment; first communication; evidence requests; respondent submissions; staff clarification; investigator and expert review; recommendations; hearing or disposition; and follow-up monitoring. Each handoff can generate rework, but not every delay is reducible through communications technology.

The intervention should be evaluated against a baseline map of an agency's real rework causes. Some additional information requests are necessary because underlying facts are disputed or because patient-safety evidence is legitimately missing. It would be inappropriate to characterize all repeat contact as inefficiency or to penalize licensees for seeking extensions, accommodations or counsel.

4. Why BoardWise is a plausible fit, not a proven agency case-management vendor

4.1 What exists today

BoardWise publicly describes a licensee-facing platform offering educational board-notice guidance, state and profession resources, structured response workflows, deadline and issue extraction, chronology and supporting-document tools, editable PDF/Word exports, form filling, signatures, hearing practice, and optional contact with independently contracted attorneys, arranged outside BoardWise and possibly at separate cost. BoardWise describes a "zero sensitive data retention" design: uploaded files stay in the user's browser, text is extracted locally, and only that extracted text is sent for processing through configured zero-data-retention services. These are vendor representations without independent certification; any public-sector use would require independent security review and verification of architecture, subprocessors, retention, logging and accessibility. [17], [20]

4.2 Why the product direction aligns with board operational pain

1. Agency vocabulary is unfamiliar to many recipients. A source-grounded educational layer can translate published procedures without replacing the notice or providing individualized legal advice.
2. Case information is often fragmented. A guided user-side chronology and exhibit checklist could make submitted materials easier for staff to identify and review.
3. Forms and deadlines vary across boards. A jurisdiction-aware rules engine is a defensible starting point if every field is reviewed against current authoritative material and easy to correct.
4. The tool begins with licensee behavior, not regulator adjudication. That creates the possibility of reducing incomplete submissions upstream without having BoardWise score the merits of a case.
5. Broader use across professions may let BoardWise adapt a common preparation layer while maintaining each agency's distinct rules, language, safeguards, and record requirements. [7], [17], [18]

4.3 What must not be represented as already established

• BoardWise has not presented independently audited, controlled evidence of fewer regulator staff hours, faster closures, improved public-safety outcomes, or lower government costs.
• A current consumer workflow should not be described as a government-certified case-management system, an official filing portal, or a substitute for a board's system of record.
• "Zero sensitive data retention," encryption, and other claims require architecture-specific verification; exported files, mail delivery, optional counsel workflows, user devices, or future integrations may have different handling characteristics.
• Anonymized anecdotes about favorable licensing outcomes do not demonstrate improved agency efficiency or causation. [17], [18]

5. Four actionable partnership models

Model A: Independent, nonexclusive public education (lowest friction)

Publish a board-specific landing page containing only official processes, links, plain-language definitions, and neutral checklists. An agency may review it for accuracy or choose whether to refer to it; the resource must explicitly state it is neither official nor endorsed unless authorized. No government systems or confidential case data are connected. All users retain access to the board's own free instructions. This is the preferred initial model. [18]

Model B: Board-reviewed packet readiness checklist

Co-design a narrow checklist for a single notice category, such as CE audit documentation or a routine information request. The user can privately assemble a chronology, identify missing attachments and create a labeled index. A checklist should flag possible incompleteness, not certify legal sufficiency. The user alone decides what to submit, and the official board intake channel remains authoritative.

Model C: Public digital-service content and accessibility testing

With appropriate authorization, test plain-language explanations, navigation, reading level, translations reviewed by qualified humans, and assistive-technology compatibility. The Department of Justice's state and local web-accessibility rule requires WCAG 2.1 AA conformance; an interim final rule effective April 20, 2026 extended compliance dates to April 26, 2027 for entities serving 50,000 or more people and April 26, 2028 for smaller entities and special district governments. Agencies should verify how the rule applies to their content and contractor arrangements. [15]

Model D: Deeper workflow integration (only after proof and procurement)

An optional later phase might use a limited, reviewed, opt-in submission manifest or status-information interface, if the regulator has a lawful procurement path, interoperable systems, robust security controls and a documented data agreement. No agency-facing feature should pull commercial BoardWise case histories into an enforcement system or provide a privileged path to outcomes. The agency retains control of receipt, service, deadlines, eligibility and records. This model is not offered or active today.

6. An eight-to-twelve-week measured pilot

Eligibility, design and control

Select one agency, one routine process and a defined set of notice types. Candidate workflows include CE audit response documentation, nonurgent document production, license-renewal disclosure instructions, or a neutral investigation process explainer. Exclude emergency matters, imminent-harm allegations, contested hearings, privileged submissions and any case where a tool could pressure an individual to waive rights. Provide an equal-quality non-digital path and keep counsel and language-access options visible.

1. Weeks 1-2, Baseline: review 30-60 anonymized historical packets and staff process steps, with agency permission; record frequencies and minutes spent on avoidable clarification, not merits. Validate rules and checklists with staff and counsel.
2. Weeks 3-4, Content test: publish a read-only educational prototype using agency-approved public sources. Run accessibility, legal-accuracy, adverse-scenario and plain-language testing with reviewers of differing familiarity.
3. Weeks 5-8, Limited rollout: offer the tool as optional to a defined cohort. Prefer randomized invitation or a matched contemporaneous comparison where ethically and operationally possible; track opt-in bias and baseline differences.
4. Weeks 9-10, Analysis: compare outcomes by notice type and complexity; audit erroneous guidance, complaints, omissions, user burden, and any disproportionate impact on individuals using accommodations.
5. Weeks 11-12, Decision: agency staff review results. Expand, revise or stop under predefined criteria. Publish an appropriately aggregated summary; do not publish confidential case details.

Pilot scorecard and decision thresholds

Illustrative workload calculation, not an impact claim

Suppose an agency receives 1,000 eligible packets annually, 30% require a follow-up solely to obtain a standard missing attachment, and each follow-up consumes 30 minutes of clerical time. If a rigorously measured pilot reduced that specific rate to 20%, estimated time avoided would be:

7. Public protection, due process and independence: non-negotiable safeguards

No influence on enforcement or adjudication

The board determines priorities, investigates facts, decides legal sufficiency, handles appeals and imposes discipline under applicable law. BoardWise must not assess credibility, predict guilt, prioritize sanctions, score practitioners for agency use or suppress material allegations. Public-safety referrals and mandatory reporting obligations remain with authorized persons and institutions. No commercial referral payment, preferential handling, co-branding or data access may create even an appearance that paying a vendor affects regulatory treatment. [18], [21]

Separate the commercial and public-interest relationships

BoardWise also sells licensee-facing services and helps users contact independently contracted attorneys, whose services are arranged outside BoardWise and may carry separate costs; BoardWise does not transmit attorney review. That creates a potential perceived or actual conflict if a regulator collaborates with the same company. A partnership protocol should explicitly prohibit regulator access to identifiable BoardWise customer information without a separate lawful basis, forbid use of attorney network data in enforcement, prohibit BoardWise from advising on individual agency cases, and identify who may authorize content changes. The regulator should have a right to withdraw references and demand prompt correction of inaccuracies. [17], [18], [19]

Transparency and rights

• Never present BoardWise as required to use, officially approved, or a preferred paid channel without lawful, explicit authorization.
• Prominently link official notices, rules, free instructions, independent legal counsel resources, and agency accommodations and language services.
• No AI-only decision-making for enforcement, discipline or procedural rights; human agency review remains available.
• Do not suggest that a checklist guarantees a complete response, deadlines are automatically extended, or the tool provides legal representation.
• Conduct accessibility, translation and legal review on documents used in official communications. [12], [13], [15]

8. Data protection, records and AI governance

A pilot should adopt the NIST AI Risk Management Framework functions (Govern, Map, Measure and Manage) and the generative-AI risk profile for hallucination, data leakage, automation bias and misuse. The U.S. GAO reported that generative AI use cases at 11 selected federal agencies increased about nine-fold, from 32 to 282, between 2023 and 2024, while agencies reported challenges including complying with federal policy and maintaining up-to-date appropriate use policies. These sources support structured risk review; they do not certify any particular BoardWise deployment. [12], [13], [14], [22]

NARA's August 21, 2026 guidance (AC 11.2026) explains when inputs, outputs, audit trails and other AI materials may be federal records and that their disposition must follow approved records schedules. It governs federal agencies, not automatically state licensing boards; it is nevertheless a useful prompt to establish applicable state records rules before storing or exchanging material. BoardWise's vendor-described approach, in which uploads and text extraction stay in the user's browser and only extracted text goes through configured zero-data-retention processing, may lower exposure for some workflows but must not be equated with exemption from public records, privacy, cybersecurity or procurement duties. [16], [17]

9. Why this aligns with 2026 public-sector modernization priorities

In NASCIO's 2026 State CIO Top 10 priorities, AI ranked first, with cybersecurity second, cost control third, modernization fourth, digital services fifth, and accessibility sixth. NASCIO's 2026 State CIO Survey reported that 98% of states have implemented enterprise AI policies and procedures. (The supplied report also cited 78% for an adopted AI governance framework; that figure could not be confirmed in the primary source and is withheld.) This suggests agencies will rightly expect a defensible business case, risk review, procurement compatibility and accessibility, not merely a compelling AI demonstration. [11]

For decision-makers, the credible pitch is a narrow service design experiment: can clearer, neutral instructions and user-side structure lower identifiable administrative costs while keeping government authority, legal rights and case information protected? That is more achievable and more defensible than presenting generative AI as a turnkey replacement for agency enforcement software.

10. Agency selection and deployment roadmap

Selection criteria for first partners

• Visible public need: documented rework or communication bottlenecks, not only a headline backlog.
• Governance readiness: an executive sponsor and reviewer from operations, legal, IT/security, records, and accessibility.
• Narrow scope: one notice class with stable written instructions and objective completeness criteria.
• Available baseline: permitted aggregated or de-identified historical counts and staff process observations.
• Independence: willingness to approve nonexclusive, non-endorsing public guidance and keep commercial transactions separate.
• Operational fit: ability to run a limited, optional trial without replacing the board's official system of record.

Roadmap to a scalable, agency-safe product

Phase 1, Public-source accuracy layer: structured policy registry, source change log, content reviewer console and a repeatable jurisdictional audit.

Phase 2, Licensee packet readiness: objective checklists, structured indexes, accessibility and safer exports; measure completeness.

Phase 3, Optional government-facing components: accessible status explanation or low-data submission manifest only where a board procures and authorizes it.

Phase 4, Shared learnings: publish aggregate patterns and a multi-board evaluation framework, never confidential cases or identifiable customer activity.

The most important product investment is likely not more autonomous reasoning but high-quality regulatory content operations: rule versioning, exception handling, document standards, jurisdiction owner review, transparent source provenance and user experience for anxious or unrepresented people.

11. Risks, failure modes and reasons not to deploy

12. Conclusions and invitation to evaluate

The public record supports three propositions. First, workload and delay are not abstractions: multiple state agencies have documented high open inventories, underperforming timeliness measures or rapidly growing investigation pipelines. Second, the workload has distinct components, some of which depend on investigative resources and professional judgment, and some of which may be improved through clearer communication and more complete submissions. Third, agency adoption of AI now takes place in a policy environment demanding measurable value, human governance, accessibility, security and independence. [1], [2], [3], [4], [5], [6], [11], [12]

BoardWise is plausibly well positioned to test the second category because its current focus is helping licensees understand processes and organize information before it reaches the agency. Its proximity to that moment of confusion is a product advantage, not an entitlement to regulatory access, evidence of efficacy, or justification for influencing outcomes. The appropriate next step is a carefully bounded evaluation with public-interest guardrails and published measurement methods. [17], [18]

Appendix A: Sample pilot charter (for agency customization)

Purpose: Evaluate whether optional, public-source-based instructions and a respondent-controlled document checklist improve first-pass administrative completeness for one defined licensing process.

Scope: 8-12 weeks, one or two notification templates, no confidential case ingestion, no automated merits determination, no automatic deadline calculation, and no agency preference for paid users.

Parties: Agency sponsor; agency legal/counsel review; operations process owner; information security/privacy; accessibility lead; BoardWise product and content owner; independent evaluation contact.

Approval gates: procurement/ethics review; written source registry; accessibility sign-off; data protection/records analysis; simulated case testing; mutually agreed stop conditions.

Primary endpoints: first-pass completeness and avoidable clerical handling time.

Secondary endpoints: clarifications, user comprehension, opt-out, accessibility issues, errors and complaints.

Evaluation: contemporaneous control or staged rollout, pre-specified case complexity adjustment, transparent limitations. Success/failure review made by agency, not vendor.

Appendix B: Suggested 10-question agency discovery interview

1. How does the board classify intake, complaints, investigations, and closure, and where are counts published?
2. Which three notice types generate the most preventable questions or missing-document follow-ups?
3. How many staff minutes are spent on administrative completeness versus evidence and substantive review?
4. Which missing items are common but do not involve discretion or factual disputes?
5. What public source controls every instruction: statute, regulation, board order, form, policy, or letter?
6. Who owns updates when the agency changes a notice template or regulation?
7. What accommodations and alternate channels are available to licensees?
8. Would an optional external educational resource require procurement, review, or a disclaimer?
9. What records, cybersecurity, privacy, conflict and ethics approvals would be prerequisites?
10. What results would justify continuation, and what findings would immediately halt the pilot?

References and source notes

Links were checked on November 1, 2026; a reachable link does not by itself establish a claim, and figures were rechecked against the cited documents as described in the editorial review below. State statistical snapshots are preserved in their source-defined periods. BoardWise sources [17]-[20] describe the company's own representations and are not independent performance evaluations.

1. [1] Arizona Auditor General, Arizona State Board of Nursing: Performance Audit and Sunset Review, Report 25-110 (September 2025). Open original source
2. [2] California Board of Registered Nursing, August 2025 Board meeting, item 4.2, Enforcement Division update (Table A, Complaint Intake). Open original source
3. [3] California Department of Finance, Board of Registered Nursing Investigations Budget Change Proposal 1111-034-BCP-2026-GB, FY 2026-27 (submitted January 10, 2026). Open original source
4. [4] Maryland Board of Nursing, Fiscal Year 2025 Annual Report. Open original source
5. [5] Maryland Department of Health, Board of Nursing Update to Health Occupations Subcommittee (January 27, 2026). Open original source
6. [6] Oregon Board of Psychology, Annual Performance Progress Report, Reporting Year 2025. Open original source
7. [7] National Council of State Boards of Nursing, 2025 Discipline Survey (generated March 30, 2026). Open original source
8. [8] NCSBN, Member Board Profiles database. Open original source
9. [9] Federation of State Medical Boards, Physician Census: 2024 census figures. (Listed in the supplied register; not cited in the report text.) Open original source
10. [10] NCSBN, Nursing Licensure Statistics. (Listed in the supplied register; not cited in the report text.) Open original source
11. [11] National Association of State Chief Information Officers, 2026 State CIO Survey. Open original source
12. [12] NIST, Artificial Intelligence Risk Management Framework (AI RMF 1.0, 2023). Open original source
13. [13] NIST, Generative AI Profile (NIST AI 600-1, 2024). Open original source
14. [14] U.S. Government Accountability Office, Generative AI Use and Management at Federal Agencies (GAO-25-107653, July 2025). Open original source
15. [15] U.S. Department of Justice, State and Local Government Web Accessibility: Small Entity Compliance Guide. Open original source
16. [16] National Archives and Records Administration, AC 11.2026, guidance on applying the Federal Records Act to AI materials. Open original source
17. [17] BoardWise, public product website and feature descriptions (vendor representations). Open original source
18. [18] BoardWise, Regulatory Board Collaboration (vendor statements of intended collaboration boundaries). Open original source
19. [19] BoardWise, Partner With BoardWise (vendor partnership positioning). Open original source
20. [20] BoardWise, About BoardWise (vendor team and product background). Open original source
21. [21] Federal Trade Commission, Active Supervision of State Regulatory Boards Controlled by Market Participants (2015). Open original source
22. [22] NIST, AI RMF Playbook: Govern, Map, Measure, Manage. Open original source

Editorial review and source check

This is BoardWise-authored analysis, not an independently verified or regulator-endorsed assessment, and not a measured BoardWise impact claim. Before publication on November 1, 2026, BoardWise editors rechecked the report's figures against the cited primary sources. The workload calculation in Section 6 is hypothetical. No pilot is active, no agency has endorsed BoardWise, and BoardWise does not ingest confidential agency case files, give agencies access to customer information, calculate deadlines automatically, or decide any matter. This article is education, not legal advice.

• Arizona: the snapshot date is January 31, 2025 per the audit's Figure 1; the 639-day average applies to all open complaints, and the 51 high-risk complaints are within the 1,543 open more than 180 days.
• California: 2,060 is the "Complaints Pending" row of the complaint-intake table, not total pending across categories. 8,317 is "Total Received" (5,330 public complaints, 1,360 convictions/arrest, 1,627 applicants). The 64% referral increase is stated as "since 2020-21." The budget proposal is a January 10, 2026 request; its enactment was not verified.
• Maryland: the FY 2025 report counts open complaints; the January 2026 briefing counts total cases (cold plus current) and includes administrative closures, so the two are not a reconciled series. The briefing describes some tracking improvements as still being implemented.
• Oregon: 18% is a 2025 reporting-year result reflecting investigations opened in 2024; the board named unexpected investigative staff absences as the most significant factor.
• NCSBN: 59 is the number of jurisdictions included in the report; 37 and 17 are the counts reporting each standard of proof for final adjudication.
• References [9] and [10] appear in the supplied register but are not cited in the text. Reference [22] in Section 8 matches the NIST AI RMF Playbook entry in the register.
• Product description corrected: the supplied text said case details are processed on the user's device. BoardWise keeps uploaded files in the browser and extracts text locally, then sends only that extracted text through configured zero-data-retention processing. No independent certification is claimed.
• Attorney contact clarified as independently contracted, arranged outside BoardWise and possibly at separate cost; BoardWise does not transmit attorney review.
• NASCIO: the 98% AI policy figure is confirmed; the 78% governance-framework figure is withheld pending primary-source confirmation. DOJ: the 2026 extension is an interim final rule effective April 20, 2026. GAO: the increase refers to reported use cases at 11 selected agencies.
• Editorial changes: em-dashes replaced with standard punctuation; the report's "verified" column label renamed "source-reported observation." Substantive discussion of AI governance in the supplied report was kept, and no AI marketing language was added.