Medical Secretaries and Administrative Assistants
43-6013.00Median wage $45,930/yr961,610 employed (US)Rank #30 of 923 scored · top 3% by substitution
Perform secretarial duties using specific knowledge of medical terminology and hospital, clinic, or laboratory procedures. Duties may include scheduling appointments, billing patients, and compiling and recording medical charts, reports, and correspondence.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution64
Exposure66
Augmentation78
Substitution — the headline: capability discounted by cost, barriers and adoption.
Exposure — technical capability alone, regardless of whether anyone deploys it.
Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.
Tasks on the substitution scale
15 rated tasks, binned by substitution score.
Position among all scored occupations
Distribution of 923 occupation scores; the marker is this occupation.
Tasks with substitution ≥ 70
20%
Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.
Why this score
The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.
Task automatabilityw 35%67
panel mean rating 3.7/5 → substitution pressure 67/100
Technical feasibility todayw 20%65
panel mean rating 3.6/5 → substitution pressure 65/100
Cost vs. human wagew 15%73
panel mean rating 3.9/5 → substitution pressure 73/100
Adoption barriersw 20%inverted — strong barriers lower the score57
panel mean rating 2.7/5 (barrier strength) → substitution pressure 57/100
Sector adoption velocityw 10%53
panel mean rating 3.1/5 → substitution pressure 53/100
Task breakdown (15 tasks)
Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.
Transmit correspondence or medical records by mail, e-mail, or fax.
77CI 67–87 · exposure 87 · augmentation 50 · importance 4.5/5 · click for rater detail
Transmit correspondence or medical records by mail, e-mail, or fax.
77| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare and medical office automation has moved rapidly; many large health systems and practices have already deployed document management and secure transmission systems, with small-practice adoption still lagging but growing steadily. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administrative functions are adopting EHR-integrated automation steadily, but many smaller practices still rely on manual fax/mail processes, giving a middling adoption pace. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Once automated, this task offers minimal augmentation opportunity because transmission itself contains no judgment or creative element requiring human-in-the-loop assistance; any residual human role is purely exception-handling. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-enabled automation tools significantly speed up sorting, routing, and confirming transmission of records while staff retain oversight for compliance and accuracy. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | This task is fully automatable end-to-end with current AI and RPA systems: document routing can be rule-based, e-mail and fax transmission are already automated by workflow software, and record handling via secure APIs meets the ≥50% time-saving threshold with zero setup overhead for most organizations. |
| Task automatability | claude-sonnet-5 | 4/5 | Transmitting documents via email/fax is largely mechanical and can be automated with document management systems, e-fax APIs, and email automation, meeting the time-saving threshold for most instances. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | HIPAA compliance and audit trails create some friction—automated systems require proper oversight and logging—but no regulatory rule mandates a human physically send records; organizations can and do use automated secure channels with appropriate controls. |
| Adoption barriers | claude-sonnet-5 | 3/5 | HIPAA and patient privacy regulations require secure, compliant transmission methods and audit trails, creating moderate friction even though the act itself isn't inherently restricted to licensed personnel. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Marginal cost of automated transmission (API calls, fax services, secure e-mail gateways) is orders of magnitude cheaper than a human secretary's loaded wage, particularly when amortized across high-volume correspondence. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated transmission via integrated software costs a fraction of a cent per transaction versus staff time spent manually mailing, emailing, or faxing records. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 5/5 | Deployed enterprise solutions (document management systems, secure fax services, automated mail handlers) reliably handle this task at scale in healthcare organizations today, with mature HIPAA-compliant platforms already in widespread production use. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | EHR systems and practice management software already automate routine record transmission (e.g., secure messaging, fax gateways, patient portals) in production at many clinics today. |
Schedule tests or procedures for patients, such as lab work or x-rays, based on physician orders.
74CI 62–85 · exposure 78 · augmentation 75 · importance 4.3/5 · click for rater detail
Schedule tests or procedures for patients, such as lab work or x-rays, based on physician orders.
74| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Healthcare is moderately digitized but slow-moving; large hospital systems are piloting scheduling automation, but widespread production deployment across primary care, smaller clinics, and rural settings remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administrative functions are adopting AI and automation tools at a moderate pace, with pilots and point solutions common but full deployment still uneven across smaller practices. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI scheduling assistants are already augmenting medical secretaries by pre-filling orders, flagging conflicts, and suggesting optimal slots; this significantly raises human productivity while keeping the secretary in a verification role. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI scheduling assistants can significantly reduce administrative burden by auto-suggesting appointment slots, sending reminders, and flagging conflicts, while staff retain oversight for exceptions and patient communication. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | Scheduling tests and procedures based on physician orders is fundamentally a data-entry and rule-matching task: receive order → check calendar → book slot → confirm. Current AI systems with access to EHRs and scheduling systems can perform this end-to-end with >50% time savings, requiring minimal human oversight. |
| Task automatability | claude-sonnet-5 | 4/5 | Scheduling tests based on structured physician orders is a rules-based coordination task well suited to AI agents that can access calendars, check availability, and confirm appointments, though occasional exceptions and multi-system integration require some human oversight. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While no hard legal requirement mandates a human schedule tests, HIPAA compliance, liability expectations around missed orders, and institutional preference for human confirmation of critical scheduling decisions create meaningful adoption friction. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists for scheduling itself, but patient privacy (HIPAA) compliance, integration with clinical systems, and need for accuracy in matching orders to correct tests create moderate organizational friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | The per-task cost of automated scheduling (API calls, small LLM inference, database queries) is orders of magnitude cheaper than medical secretary labor, even accounting for healthcare-grade infrastructure and compliance overhead. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated scheduling systems can handle high volumes of routine bookings at a fraction of the cost of a human secretary's time once integrated, though initial setup and EHR integration add cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed healthcare scheduling systems and EHR-integrated AI already handle test scheduling in many large health systems; however, integration with legacy systems and varying clinic workflows introduces material friction, keeping this below full production ubiquity. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Some healthcare scheduling software and AI-driven patient coordination tools exist in production, but many practices still rely on manual phone-based scheduling due to fragmented EHR/lab systems and interoperability issues. |
Transcribe recorded messages or practitioners' diagnoses or recommendations into patients' medical records.
73CI 67–79 · exposure 75 · augmentation 100 · importance 4.6/5 · click for rater detail
Transcribe recorded messages or practitioners' diagnoses or recommendations into patients' medical records.
73| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare organizations, especially larger health systems and hospitals, have rapidly adopted AI-powered clinical transcription and dictation tools over the past 3–5 years. Adoption is measurable and accelerating in information-intensive healthcare settings, though smaller practices lag. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare has historically been a slower-adopting sector due to compliance and legacy systems, but ambient documentation and AI scribe tools have seen rapid recent uptake, placing this at middling-to-accelerating adoption. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | Even where humans remain in the loop for review and editing, AI transcription dramatically accelerates the task by eliminating manual typing and reducing dictation-to-record time. The human secretary's productivity is substantially raised while maintaining control and oversight. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI scribes and transcription tools substantially reduce administrative burden while a human still reviews and finalizes entries into the record, representing a strong augmentation use case. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Speech-to-text AI systems can reliably transcribe medical audio with high accuracy, and current tools (like Nuance Dragon Medical, Otter.ai, and others) achieve substantial time savings at comparable quality. However, specialized medical terminology and context-dependent interpretation occasionally require human review, preventing a perfect 5. |
| Task automatability | claude-sonnet-5 | 4/5 | Medical transcription and dictation-to-note conversion is well within current ASR and LLM capabilities, especially with structured templates and speaker context, meeting the 50% time-saving bar for most straightforward dictations. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While medical records are regulated (HIPAA), the transcription task itself has no legal requirement that a human must perform it; oversight and error-checking can be human-led but the transcription itself is automatable. Primary barriers are organizational inertia and preferences for human review, not hard legal constraints. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement for transcription itself, but HIPAA compliance, EHR integration requirements, and the need for accuracy verification (since errors in medical records carry liability risk) create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI transcription costs per utterance are typically a fraction of the loaded hourly wage of a medical secretary, and no specialized licensing or equipment is required beyond standard cloud subscriptions. The cost advantage is substantial and well-established. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI transcription costs (cents to low dollars per note) are far below the loaded cost of a human medical secretary performing the same transcription task, even accounting for oversight/editing time. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple mature products (clinical transcription services, EHR-integrated speech recognition) are deployed in production medical settings and demonstrate reliable performance at scale. Some variation in accuracy across accents and audio quality exists, but the core task is demonstrably performed reliably by deployed systems. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Products like Nuance DAX, Dragon Medical, and various ambient clinical documentation tools are deployed at scale in hospitals and clinics today, though accuracy on complex terminology or noisy audio still requires human review. |
Operate office equipment, such as voice mail messaging systems, and use word processing, spreadsheet, or other software applications to prepare reports, invoices, financial statements, letters, case histories, or medical records.
70CI 67–72 · exposure 75 · augmentation 100 · importance 4.4/5 · click for rater detail
Operate office equipment, such as voice mail messaging systems, and use word processing, spreadsheet, or other software applications to prepare reports, invoices, financial statements, letters, case histories, or medical records.
70| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Healthcare and administrative sectors show uneven adoption: many large healthcare systems are piloting AI transcription and document assist tools, but smaller practices and risk-averse organizations remain slow. Pilots are common, but deep production displacement is still emerging rather than established. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administrative functions are adopting AI scribes and document tools at a moderate pace, behind finance/tech but ahead of manual labor sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI significantly augments administrative work today: real-time transcription assistance, auto-complete in medical records, template-based form filling, and report drafting all materially raise human productivity while the secretary remains the quality gate. This is transformative assistance even if not full automation. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI substantially speeds up drafting, formatting, and transcription work while the secretary retains responsibility for verification and finalization. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI can handle most of this task end-to-end: voice mail transcription (speech-to-text), document generation (word processing via LLMs), spreadsheet operations, and basic medical record formatting are all achievable with >50% time savings. The main friction is context integration and quality assurance, but the core workflow is largely automatable. |
| Task automatability | claude-sonnet-5 | 4/5 | Drafting reports, letters, invoices, and structuring case histories/medical records via templates and dictation-to-text is well within current AI capability, though some manual data entry and system navigation remains. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Healthcare regulatory requirements (HIPAA, medical record accuracy) and organizational policies mandate human review and sign-off on clinical documentation, creating material friction. However, the automation itself is not legally barred; the barrier is oversight and liability asymmetry, not licensing. |
| Adoption barriers | claude-sonnet-5 | 2/5 | Some oversight is needed for medical record accuracy and privacy (HIPAA), but no licensing requirement mandates a human perform this specific documentation task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference + integration for transcription, document generation, and spreadsheet work costs pennies per task versus an administrative assistant's loaded wage (~$30–45/hour). Even with oversight overhead, the cost ratio heavily favors automation by 5–10×. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI-assisted drafting and voice transcription tools cost a small fraction of secretarial hourly wages for equivalent document production. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products demonstrably perform these subtasks in production: transcription services (Otter, Google Docs voice typing), document generation (Word + copilot, Claude, ChatGPT), and spreadsheet automation (Excel with Copilot, Zapier) are used at scale in healthcare settings today. Reliability is high for routine tasks, though medical record compliance still requires oversight. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Mature products (EHR-integrated dictation, AI scribes, Office/Google AI assistants) already generate documents and reports in production medical office settings, though full end-to-end automation across all listed equipment types is less complete. |
Receive and route messages or documents, such as laboratory results, to appropriate staff.
69CI 60–79 · exposure 70 · augmentation 75 · importance 4.5/5 · click for rater detail
Receive and route messages or documents, such as laboratory results, to appropriate staff.
69| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare systems and practices are actively deploying automated document routing and lab result distribution, driven by EHR vendors and RPA adoption in back-office operations, with measurable displacement of clerical routing work. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administrative back-office functions lag behind sectors like finance and tech in AI adoption due to legacy systems, interoperability issues, and cautious regulatory environment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can assist by pre-sorting and flagging priority messages, suggesting routing destinations, and surfacing critical results for human review, significantly reducing manual scanning time while maintaining final human control over sensitive triage decisions. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted triage and routing tools meaningfully reduce manual sorting burden and flag urgent items, letting staff focus on exceptions and patient-facing tasks. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can reliably classify and route messages/documents (lab results, referrals) using document parsing and rule-based routing with 50%+ time savings on standard templates and recurring message types. However, edge cases requiring human judgment about priority or recipient may still need oversight. |
| Task automatability | claude-sonnet-5 | 4/5 | Routing messages and lab results based on rules or content classification is a well-structured task that AI systems (e.g., inbox triage, EHR integration bots) can handle end-to-end with significant time savings, though some edge cases need human judgment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Light regulatory friction exists around audit trails and error accountability in healthcare, but no law requires a human secretary to route lab results; most barriers are organizational (preference for human backup) rather than legal. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Handling lab results triggers HIPAA compliance and patient-safety concerns, requiring audit trails and human oversight for abnormal or urgent results, creating moderate friction despite no strict licensing requirement for routing itself. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automation cost (OCR, classification, routing rules, minimal oversight) is orders of magnitude cheaper than paying a human administrative assistant for repetitive message triage across hundreds of documents daily. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated routing via existing EHR/messaging software is very cheap per transaction compared to a human secretary's time, though initial integration and oversight add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed RPA and intelligent document processing solutions already handle message routing and triage in healthcare settings at scale, with automated lab result distribution now standard in many EHR systems. Minor gaps remain in unusual formats or ambiguous recipient determination. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | EHR systems and healthcare communication platforms have automated routing features, but many clinics still rely on manual triage due to integration gaps, formatting variability, and reliability concerns with critical results. |
Maintain medical records, technical library, or correspondence files.
69CI 67–70 · exposure 75 · augmentation 75 · importance 4.5/5 · click for rater detail
Maintain medical records, technical library, or correspondence files.
69| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare is a highly digitized sector with strong financial incentives to reduce administrative overhead. Major health systems have widely deployed electronic health record systems with integrated document management, though smaller practices and rural clinics lag. Overall adoption is solidifying from pilot to production phase. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administrative functions have adopted EHR and digital records systems substantially, but full AI-driven records management is still uneven across practice sizes and settings. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-powered search, automated tagging, and intelligent file suggestions significantly augment human secretaries' ability to locate and organize records. These systems help workers manage larger volumes and reduce manual indexing effort while the human remains in control of sensitive decisions and quality oversight. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools significantly speed up filing, retrieval, and organization of records and correspondence, letting administrative assistants focus on exceptions and patient-facing tasks. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | File maintenance tasks—organizing, categorizing, storing, and retrieving medical records and correspondence—can be substantially automated using document management systems, OCR, and AI-powered classification. Current systems can handle 70%+ of routine filing, indexing, and records organization at scale, though some manual review of sensitive or ambiguous documents may remain necessary. |
| Task automatability | claude-sonnet-5 | 4/5 | Filing, indexing, and maintaining digital records/correspondence is largely structured data management that current AI/EHR systems and document automation tools handle well, though some physical filing or edge-case categorization still needs human input. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Healthcare records automation faces moderate barriers: HIPAA compliance requirements demand careful system design and oversight, liability concerns around misfiled or lost records create organizational caution, and some healthcare systems remain tethered to legacy systems. Patient contact and signature requirements for some correspondence add friction but do not prevent the bulk of file maintenance automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | HIPAA and medical records regulations require accuracy, security, and audit trails, and organizations often keep human oversight for compliance and error correction, creating moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Once implemented, cloud-based document management and automated classification systems cost significantly less per record processed than hiring administrative staff. A single instance can handle records for hundreds of patients, easily achieving 5–10x cost advantage compared to manual filing. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated records management software and cloud storage systems cost far less per record processed than dedicated clerical labor, though initial integration and compliance setup add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed document management and records systems (e.g., Epic, Cerner, Athenahealth) with AI-assisted classification and OCR are widely used in healthcare organizations. These products reliably perform file organization, digitization, and retrieval in production environments, though integration complexity and legacy system compatibility can create friction. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | EHR systems, document management platforms, and AI-assisted indexing/retrieval tools are widely deployed in healthcare settings today for records maintenance, though full end-to-end automation without any human oversight is less common. |
Compile and record medical charts, reports, or correspondence, using typewriter or personal computer.
68CI 62–74 · exposure 70 · augmentation 100 · importance 4.5/5 · click for rater detail
Compile and record medical charts, reports, or correspondence, using typewriter or personal computer.
68| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare is a digitally mature sector with strong financial incentives to reduce administrative burden; EHR vendors and healthcare systems are rapidly deploying AI-assisted documentation tools, with measurable adoption in hospitals and large clinics observable today. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administrative functions are adopting AI documentation tools steadily but healthcare overall lags behind finance/tech in AI deployment due to compliance and legacy systems. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically amplifies administrative assistant productivity by auto-populating charts from clinical encounters, generating draft correspondence, and organizing records, allowing staff to focus on exception handling and patient interaction rather than manual data entry. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI dictation, transcription, and auto-population tools significantly speed up chart compilation and correspondence drafting while the secretary reviews and finalizes records. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Current AI systems can reliably extract, summarize, and structure medical information from voice recordings, dictations, or scanned documents into electronic charts and reports. OCR and NLP systems achieve high accuracy on medical text, and template-based document generation can produce correspondence with minimal human intervention, meeting the ≥50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 4/5 | Compiling and recording medical documentation is largely structured data entry and transcription work that current AI (speech-to-text, NLP extraction, EHR integrations) can handle with substantial time savings, though some human review remains needed for accuracy. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Medical charts and records have regulatory requirements (HIPAA, state licensing boards) and liability implications (documentation errors can affect patient care), creating oversight and compliance friction. However, AI does not legally need to be supervised by a licensed physician at the documentation stage, only reviewed, moderating the barrier. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Medical records require accuracy and compliance with HIPAA and institutional recordkeeping standards, creating moderate oversight requirements, though the compiling/recording task itself isn't a licensed clinical act. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-based medical documentation (transcription, chart compilation, correspondence generation) costs a fraction of human secretary labor per task; a single AI system serves hundreds of providers at marginal cost, easily achieving an order-of-magnitude cost advantage. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI transcription/documentation tools cost a small fraction of a secretary's hourly wage per document processed, though integration and oversight costs reduce the savings somewhat. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products (EHR systems with AI-assisted documentation, medical transcription AI, and chart automation tools) demonstrably perform this task in production healthcare settings, though they typically require human review for accuracy and liability reasons, slightly limiting the rating from 5. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Ambient clinical documentation tools and medical transcription AI are deployed in production at many practices, but error rates and need for human correction still limit full reliability across all chart types and specialties. |
Answer telephones and direct calls to appropriate staff.
67CI 59–75 · exposure 67 · augmentation 75 · importance 4.6/5 · click for rater detail
Answer telephones and direct calls to appropriate staff.
67| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare and professional services sectors are actively adopting AI call routing and answering systems in production; major EHR and phone system vendors have integrated these capabilities, and deployment is accelerating in both large systems and mid-sized clinics. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administrative functions are adopting AI call-handling tools, but overall sector digitization and adoption lag behind pure information/finance industries. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can assist by providing real-time caller information, suggested routing, decision support for complex calls, and automatic documentation of call summaries, significantly raising a secretary's handling capacity and accuracy while they remain the final decision-maker on transfers. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI significantly assists by triaging, transcribing, and pre-routing calls so staff handle fewer routine interactions and focus on complex ones. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Call routing and answering can be partially automated via IVR and AI-based call screening systems that recognize intent and transfer to the appropriate department, achieving meaningful time savings. However, complex inquiries requiring contextual judgment or escalation still require human intervention, preventing full end-to-end automation at the 50% threshold. |
| Task automatability | claude-sonnet-5 | 4/5 | AI phone agents and IVR systems with NLU can answer calls and route them to appropriate staff with high accuracy for routine requests, meeting the time-saving threshold for a large share of calls. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Moderate barriers exist: healthcare organizations often have patient privacy (HIPAA) and data security policies that require careful vendor selection and audit; some patients prefer human contact, and liability concerns around call misrouting create organizational friction but no legal prohibition. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement to answer/route calls, but medical contexts carry some liability and privacy (HIPAA) concerns plus patient preference for human contact in sensitive situations. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-based call answering and routing (via cloud platforms with modest setup) costs significantly less per call than a full-time secretary's loaded wage, with inference costs of cents per call versus $25-30/hour for human staff. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Voice AI/IVR subscription costs are far lower than a receptionist's loaded wage for handling routine call volume, though integration and oversight add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products like AI call centers, voice assistants (Amazon Connect, Google Cloud Contact Center AI), and intelligent call routing systems are in production use today across many healthcare and administrative settings, though some edge cases and accent/dialect challenges remain. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Deployed AI call-answering and routing products (e.g., healthcare-focused conversational IVR/voice AI) are used in production at medical practices today, though complex or sensitive calls still often escalate to humans. |
Schedule and confirm patient diagnostic appointments, surgeries, or medical consultations.
67CI 59–75 · exposure 67 · augmentation 75 · importance 4.5/5 · click for rater detail
Schedule and confirm patient diagnostic appointments, surgeries, or medical consultations.
67| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare organizations increasingly adopt EHR scheduling automation and patient-facing booking portals; large health systems and private practices show active deployment, though adoption remains uneven across smaller/rural practices and specialty centers with complex scheduling needs. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administration is adopting automated scheduling and reminder systems at a moderate pace, with common deployment for routine confirmations but slower uptake for complex scheduling due to EHR integration challenges and regulatory caution. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-driven scheduling assistants (smart availability matching, automated reminders, patient intake pre-filling) substantially improve human scheduler productivity by reducing manual calendar cross-checking and follow-up calls, allowing staff to focus on complex requests and patient communication. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI scheduling tools substantially assist administrative staff by automating reminders, confirmations, and calendar management while humans still handle exceptions and complex cases. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Scheduling and confirmation can be partially automated via calendar systems and patient communication APIs, but handling complex cases (doctor preferences, insurance verification, patient availability across multiple time zones, last-minute changes, and conflict resolution) typically requires human judgment and empathy, limiting time savings below 50%. |
| Task automatability | claude-sonnet-5 | 4/5 | Scheduling and confirming appointments is a structured, rules-based task involving calendar logic, availability matching, and templated communication, which AI scheduling agents can already handle for a large share of cases with significant time savings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | HIPAA compliance, patient privacy requirements, and organizational integration with clinical staff create meaningful friction; however, no legal requirement mandates a human perform scheduling, and many practices already use partial automation without regulatory obstruction. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement exists for scheduling itself, though HIPAA-related data handling and patient preference for human contact for sensitive scheduling (e.g., surgery) create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Integration of scheduling automation into existing EHR infrastructure is already amortized across many clinical workflows; marginal cost per appointment is low relative to hourly medical secretary wages (typically $30–$45k annually), making automation economically attractive. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated scheduling and confirmation via software/chatbots costs a small fraction of a human secretary's wage per interaction, though integration with legacy medical systems adds some overhead. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed products (EHR systems like Epic, Cerner; scheduling platforms like Acuity, Calendly integrations) routinely handle appointment scheduling in production, though many still require manual confirmation calls and human oversight to manage edge cases and patient communication nuances. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Deployed products (e.g., AI scheduling assistants integrated with EHR/practice management systems, automated reminder/confirmation systems) already handle appointment scheduling and confirmation in many clinics, though exceptions and complex multi-provider surgical scheduling still require human intervention. |
Prepare correspondence or assist physicians or medical scientists with preparation of reports, speeches, articles, or conference proceedings.
67CI 59–75 · exposure 67 · augmentation 100 · importance 3.7/5 · click for rater detail
Prepare correspondence or assist physicians or medical scientists with preparation of reports, speeches, articles, or conference proceedings.
67| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare and medical publishing sectors have rapid pilot and early-production adoption of AI writing assistants, especially in administrative and academic medical centers; digital workflows and high literacy enable faster deployment than laggard sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administration is adopting AI writing/documentation tools steadily but is generally slower than finance or tech due to compliance concerns, EHR integration complexity, and conservative IT practices. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI substantially assists physicians and secretaries by generating first drafts, auto-formatting, and suggesting edits for reports and correspondence, materially raising productivity while physicians retain final decision-making and clinical accuracy responsibility. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is highly effective as a drafting and editing assistant for correspondence, reports, and speeches, letting the human focus on review, accuracy, and personalization. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can automate significant parts of drafting, formatting, and organizing medical correspondence and reports, but cannot fully replace the task without physician review and approval. Medical accuracy, liability, and context-specific clinical details typically require human oversight, limiting end-to-end automation to roughly 50% time savings. |
| Task automatability | claude-sonnet-5 | 4/5 | Drafting correspondence, reports, and speeches from notes or bullet points is well within current LLM capability, requiring only human review and light editing to meet quality bar. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Physicians and medical institutions retain discretion and liability for final documents, requiring human review and sign-off; professional standards and institutional policies often mandate physician oversight of patient-facing or publication correspondence, creating moderate friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for drafting correspondence or reports, though physician review/sign-off is expected for medical accuracy and liability reasons, creating modest friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference cost for drafting and formatting is minimal (fractions of a dollar per task), compared to a secretary's loaded wage (typically $25–40/hour). Oversight costs are low, making AI substantially cheaper on a per-output basis. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI drafting tools cost a fraction of a secretary's hourly wage for equivalent word-processing output, though integration and review time reduce savings somewhat. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature LLM products and medical writing assistants (e.g., GPT-4, specialized medical AI) demonstrably draft medical correspondence and reports in production settings with good accuracy on routine content, though error rates remain non-negligible on complex clinical terminology or novel findings. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Deployed products like Microsoft Copilot, Google Workspace AI, and dictation/transcription-to-draft tools are already used in medical offices for correspondence and document drafting, though medical accuracy checks remain necessary. |
Complete insurance or other claim forms.
65CI 62–67 · exposure 70 · augmentation 75 · importance 4.5/5 · click for rater detail
Complete insurance or other claim forms.
65| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Healthcare organizations have adopted RPA and claim automation in pilots and early production, but rollout remains patchy. Larger health systems and insurance processors move faster; smaller practices lag. Overall adoption is middling—common in forward-thinking organizations, less so across the sector. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administration is adopting automation steadily via EHR-integrated billing tools, but overall healthcare sector digitization lags behind finance or tech, with many small practices still manual. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI assists medical secretaries by auto-populating forms from structured records, identifying missing data, and flagging likely errors, which significantly raises human productivity even when the human retains final review and submission responsibility. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted coding, form auto-population, and error-checking tools significantly speed up secretaries' claim completion while they retain final review and submission responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Insurance claim forms are highly structured documents with standard fields, data entry requirements, and predictable logic. AI can extract relevant information from medical records and patient data, populate forms accurately, and route them automatically with minimal human intervention, achieving well over 50% time savings at comparable quality. |
| Task automatability | claude-sonnet-5 | 4/5 | Filling out structured insurance/claim forms from EHR data and coding is a well-defined, template-driven task that current AI plus RPA tools can largely handle, though edge cases and payer-specific rules require oversight. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While no law explicitly prohibits AI form completion, regulatory oversight (HIPAA compliance, insurance regulations), organizational risk-aversion, and insurance company acceptance policies create moderate friction. Many healthcare organizations still require a human to review or sign off on claims before submission. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement to complete forms, but billing errors carry financial and compliance risk (HIPAA, insurance fraud liability), so organizations often keep human sign-off in the loop. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-based form automation (inference plus integration) costs substantially less than hiring administrative staff per completed form. Even accounting for oversight and exception handling, the cost per claim is likely 5–10× cheaper than the loaded wage of a medical secretary. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated claims processing software is dramatically cheaper per claim than manual entry by a paid secretary, though initial integration with practice management systems adds some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed products (RPA platforms, healthcare-specific form automation, document processing AI) are performing insurance claim form completion in real healthcare organizations. Some edge cases and complex scenarios require human review, but mainstream scenarios are handled reliably at scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Medical billing software and claims clearinghouses already automate much of this with AI-assisted coding suggestions, but denial rates and error correction still require human review in most practices. |
Perform bookkeeping duties, such as credits or collections, preparing and sending financial statements or bills, and keeping financial records.
64CI 50–79 · exposure 62 · augmentation 75 · importance 4.2/5 · click for rater detail
Perform bookkeeping duties, such as credits or collections, preparing and sending financial statements or bills, and keeping financial records.
64| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare, finance, professional services, and most mid-to-large organizations have already adopted accounting software with significant automation; adoption is rapid and widespread in digitized sectors, though small practices lag behind. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Medical offices have moderately adopted billing/EHR-integrated financial software, but small practices often lag and full automation of collections remains limited. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI augments secretaries by auto-categorizing transactions, flagging anomalies, auto-populating forms, and generating draft statements, substantially raising productivity even when humans retain control and review authority. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-powered billing and accounting software significantly speeds up statement generation, payment tracking, and record organization, meaningfully boosting a secretary's productivity while they retain oversight. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | This task involves routine data entry, record-keeping, and financial document generation—all highly automatable with current systems. RPA, accounting software, and AI agents can handle credits, collections tracking, statement generation, and record maintenance with minimal human intervention, easily meeting the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 3/5 | AI/software can automate invoice generation, billing statements, and record-keeping with accounting integrations, but collections follow-up and exception handling still require human judgment and setup effort. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Moderate barriers exist: audit and compliance requirements often mandate human review and sign-off of financial records, and some organizations prefer human oversight for fraud detection and exception handling. However, no strict legal prohibition prevents automation of the core bookkeeping work itself. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement blocks automated bookkeeping, though healthcare billing involves HIPAA compliance and financial accuracy concerns that create moderate organizational caution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-powered accounting systems cost a fraction of a full-time secretary's loaded wage once amortized across an organization, often $10–50/month per user versus $40,000–60,000+ annual salary, making the cost ratio highly favorable. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Billing software subscriptions are cheaper than a dedicated bookkeeper's time for routine tasks, but human review for discrepancies, insurance issues, and collections keeps overall cost comparable rather than order-of-magnitude lower. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature accounting and bookkeeping software (QuickBooks, SAP, NetSuite) with AI-driven capabilities already perform these functions reliably in production across many organizations. Errors occur mainly with edge cases or unusual transactions, but core bookkeeping is demonstrably automated at scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Practice management and billing software (with automated invoicing, payment reminders) is widely deployed in medical offices, but full bookkeeping including collections and reconciliation still typically needs human oversight and correction. |
Perform various clerical or administrative functions, such as ordering and maintaining an inventory of supplies.
62CI 52–72 · exposure 62 · augmentation 75 · importance 3.6/5 · click for rater detail
Perform various clerical or administrative functions, such as ordering and maintaining an inventory of supplies.
62| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Healthcare has moderate adoption of supply-chain automation, with large health systems deploying advanced procurement systems but smaller practices and clinics still relying on manual ordering. Adoption is faster in information-heavy sectors but slower in fragmented medical settings. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Small medical practices and administrative support roles are generally slower adopters of advanced automation tools compared to purely digital-native sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI strongly augments secretaries' inventory work by generating alerts, consolidating vendor prices, drafting purchase orders, and highlighting discrepancies for human review—allowing them to focus on exception handling and vendor relationship management rather than routine data entry. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-enabled inventory and procurement tools can significantly streamline tracking, forecasting reorder needs, and reducing manual counting work for administrative staff. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Ordering and inventory maintenance is highly automatable: AI can track stock levels, predict reorder points, place orders via APIs, and reconcile receipts against purchased items. Current systems achieve >50% time savings with integration to procurement platforms, though final human approval on large orders may still be needed. |
| Task automatability | claude-sonnet-5 | 3/5 | Inventory tracking and reordering can be substantially automated via software (thresholds, auto-reorder), but physical receiving, verifying deliveries, and vendor exceptions still require human involvement. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Barriers are relatively low: no licensing requirement to automate ordering, minimal legal liability for supply ordering errors, and organizations have wide discretion in automating clerical tasks. Customer contact preference and organizational change management are modest friction points. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for ordering supplies, though organizational trust and established vendor relationships create some friction to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-driven inventory automation costs (software licensing, setup, API integration) are substantially lower than the annual loaded cost of a medical secretary's time spent on ordering and stock-tracking, typically yielding 3–10× cost savings once deployed. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Inventory software subscriptions are cheap relative to labor, but integration with existing practice management systems and ongoing human oversight keep costs roughly comparable rather than dramatically lower. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature inventory management and procurement automation systems are deployed in hospitals and medical practices today (e.g., ERP modules, supply-chain APIs, automated reordering). These perform reliably at scale for routine consumables, though edge cases and exceptions still require oversight. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Inventory management systems and procurement software are widely deployed in healthcare offices, but full automation of ordering decisions and supply verification still involves human oversight in most practices. |
Interview patients to complete documents, case histories, or forms, such as intake or insurance forms.
46CI 34–59 · exposure 45 · augmentation 75 · importance 4.4/5 · click for rater detail
Interview patients to complete documents, case histories, or forms, such as intake or insurance forms.
46| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare remains cautious and highly regulated; adoption of AI for patient-facing intake is in pilot phase at most organizations, with slow enterprise roll-out. Most medical offices still rely on manual secretary-conducted interviews. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administration is adopting patient-facing AI intake tools steadily but unevenly, with many practices still using paper or basic digital forms rather than full AI-driven interviews. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can effectively assist secretaries by auto-populating forms from patient responses, flagging missing fields, and suggesting corrections in real time, meaningfully raising human productivity while the secretary retains control and quality oversight. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can pre-fill forms, transcribe conversations, and flag missing information, significantly speeding up the administrative assistant's workflow while they retain oversight of accuracy and patient rapport. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can automate partial intake workflows (form filling from spoken or written input, basic data extraction), but patient interviews require nuanced communication, follow-up clarification, and handling of sensitive health information that current systems struggle with reliably. Rough 40-60% time savings is achievable on structured portions, not the full task. |
| Task automatability | claude-sonnet-5 | 3/5 | AI voice agents and chatbots can conduct structured intake interviews and populate forms, but handling ambiguous patient responses, sensitive medical history nuance, and edge cases still requires human oversight for full accuracy. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | HIPAA regulations, liability for missing or incorrect patient information, quality assurance requirements, and institutional preference for human verification create strong adoption friction. Errors in intake can cascade through clinical care, raising error-cost asymmetry. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for this specific task, but privacy regulations (HIPAA), data accuracy liability, and patient preference for human interaction during health disclosures create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI inference and integration costs for healthcare-compliant systems, plus mandatory human review and error correction, approach or exceed the hourly wage of a medical secretary. Total cost-in is not yet favorable at scale. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated intake forms and voice/chat agents cost far less per patient interaction than staff time, though integration with EHRs and insurance systems adds some overhead. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While chatbots and voice AI exist for basic intake, no mature production system reliably handles the full interview-to-completed-form pipeline with acceptable error rates for healthcare compliance. Most deployments are narrow pilots requiring heavy human oversight. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Digital intake platforms and conversational AI tools exist in production at some clinics (patient self-service kiosks, chatbot intake), but many practices still rely on staff-conducted interviews due to accuracy and patient comfort concerns. |
Greet visitors, ascertain purpose of visit, and direct them to appropriate staff.
20CI 10–30 · exposure 17 · augmentation 50 · importance 4.5/5 · click for rater detail
Greet visitors, ascertain purpose of visit, and direct them to appropriate staff.
20| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Medical offices—particularly small-to-medium practices that employ most medical secretaries—lag in automation adoption and place high cultural value on human greeting and triage, limiting actual deployment of AI reception systems. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administrative front-desk functions are digitizing slowly compared to other office functions; kiosks are common in some large systems but far from universal, especially in smaller practices. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist by suggesting staff routing based on visitor descriptions, queuing appointments in real-time, or flagging common visitor types, but the human receptionist remains essential for the social and judgment-based core of the task. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Check-in software, scheduling systems, and automated queuing can assist staff by pre-processing visitor information and directing routine cases, improving efficiency while humans still manage exceptions. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires real-time human interaction, social judgment about visitor intent, and contextual knowledge of staff availability and office layout. Current AI cannot reliably execute the full greeting-to-direction workflow in physical spaces without constant supervision. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical greeting and directing of in-person visitors requires presence and situational judgment (e.g., handling distressed patients, security concerns) that current AI cannot fully replicate, though check-in kiosks partially cover this. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Medical offices have strong preferences for human-first visitor interactions for liability and patient experience reasons; HIPAA compliance concerns around who/what handles visitor information; and organizational friction from the loss of a receptionist role that often includes other critical functions. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement, but healthcare settings favor human contact for triage-adjacent interactions, privacy (HIPAA) concerns with automated systems, and patient comfort with human staff create meaningful friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | An automated reception system (physical kiosk or AI agent) would require significant hardware, software, and integration costs that would exceed the loaded wage of a medical secretary for this task, especially given the need for fallback human support. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Kiosk/tablet systems have upfront and maintenance costs and still require staff backup for exceptions, so total cost savings versus a receptionist are moderate at best. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While chatbots can handle simple greeting scripts in digital channels, no deployed product reliably greets physical visitors, assesses their needs through conversation, and directs them to staff in real-world medical office settings at scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Self-service kiosks and digital check-in exist in many clinics, but they handle routine registration rather than the full nuanced task of ascertaining purpose and directing visitors, especially exceptions or urgent cases. |
Related occupations — Office & Administrative Support
How to read this
A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.
What would change this score
New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.