Name role, authority, jurisdiction, and task
Define one bounded job, people/rights/money/property affected, approved systems, and the qualified reviewer.
Occupation directory / Evidence before automation
A static shelf of completed StackPilot occupation manuals and family guide shops, with evidence artifacts, approval boundaries, and planned coverage kept visibly separate.
Short answer: choose the manual matching the role and exact task. Each specialist route shows what AI may prepare, what evidence must remain visible, and which decisions or external actions stay with the licensed, qualified, authorized human.
Start Here bridge status / no data collected
Every listed occupation page includes a static bridge back to the buyer → pain → offer → stack → agents → content → customers → delivery order, so specialist visitors can slow down before tools, agents, regulated decisions, or customer-facing actions.
No forms · no capture · no checkout · no outreachBefore a specialist manual
Occupation manuals are not a shortcut around business clarity. If the reader is still choosing who to help or what to sell, route them back through buyer, pain, offer, stack, agents, content, customers, and delivery before any profession-specific automation.
Development, QA, data warehousing, websites, apps, agents, automation, credentials, and file routes.
Open shop →Real estateRealtors, appraisers, property/community managers, mortgage operators, new-home teams, and supporting task routes.
Open shop →InsuranceLicensed agent/client-review work and claims/policy-service preparation with distinct authorities.
Open shop →Banking & lendingLoan officers, intake clerks, creditors, valuation, compliance, closing, security, and funding routes.
Open shop →A safe AI workflow for accounting source control, reconciliation preparation, variance questions, close checklists, private financial data, and human review.
Open manual →Occupation manualA source-bound AI workflow for claims evidence review, unknown logs, reviewer questions, and authorized human handoffs.
Open manual →Occupation manualA proof-first AI workflow for clinical data provenance, controlled study baselines, neutral query drafts, validation evidence, privacy, audit trails, and lock review.
Open manual →Occupation manualA proof-first AI workflow for network-change evidence, validation planning, and rollback handoffs without delegating access, design, execution, or incident decisions.
Open manual →Occupation manualPrepare a source-bound network-support review packet with AI while keeping network access, logs, diagnosis, severity, response, configuration, communication, and system actions human-controlled.
Open manual →Occupation manualPrepare a source-bound systems-analysis review packet with AI while system access, requirements, scope, design, tests, change, communication, and external actions remain human-controlled.
Open manual →Occupation manualA proof-first AI workflow for system requirements, architecture alternatives, interfaces, threat questions, ADR drafts, review routes, and approval-ready handoffs.
Open manual →Occupation manualA review-only field manual for preparing source-bound end-user support intake packets with clear authority, uncertainty, and no-action boundaries.
Open manual →Occupation manualA review-only AI workflow for credit analysts: prepare source-bound evidence packets while keeping private data, credit decisions, notices, and communications human-controlled.
Open manual →Occupation manualUse AI to prepare a neutral credit-counseling intake handoff without giving debt, credit, bankruptcy, or legal advice.
Open manual →Occupation manualA proof-first AI workflow for source-bound support drafts, missing-evidence flags, customer-data boundaries, escalations, and human-reviewed handoffs.
Open manual →Occupation manualA proof-first AI workflow for source contracts, field lineage, data validation, reconciliation, quality metrics, query cost, incidents, and warehouse change review.
Open manual →Occupation manualA proof-first AI workflow for database inventory, least privilege, backup/restore evidence, migrations, performance diagnosis, integrity, and recovery review.
Open manual →Occupation manualA source-bound AI workflow for document-control evidence, version labels, unknown logs, and authorized human handoffs.
Open manual →Occupation manualA practical, evidence-first AI workflow for quantitative financial research: scoped inputs, reproducibility, limitations, human challenge, and approval-ready handoffs.
Open manual →Occupation manualA source-bound AI workflow for security-signal evidence, unknown logs, and authorized human review handoffs.
Open manual →Occupation manualA source-bound AI workflow for security-engineering labels, unanswered questions, and authorized human review—without assessments, control decisions, tests, changes, or deployment.
Open manual →Occupation manualA safe AI workflow for insurance intake, source-linked policy facts, renewal difference review, coverage questions, private data, and licensed human approval.
Open manual →Occupation manualA source-bound AI workflow for auto-damage evidence inventories, unknown logs, review questions, and authorized estimate-review handoffs.
Open manual →Occupation manualA proof-first AI workflow for insurance claim intake, policy-service evidence, identity gates, document versions, correspondence QA, escalation, and audit trails.
Open manual →Occupation manualA source-bound AI workflow for insurance-underwriting evidence inventories, unknown logs, review questions, and authorized human handoffs.
Open manual →Occupation manualA proof-first AI workflow for loan application intake, field provenance, document evidence, fair-lending safeguards, FCRA boundaries, milestones, and human handoffs.
Open manual →Occupation manualAn evidence-first AI workflow for customer research, source ledgers, positioning hypotheses, claim review, campaign experiments, and measurement.
Open manual →Occupation manualA safe AI workflow for mortgage document intake, missing-item follow-up, borrower question prep, disclosure tracking, data protection, and human credit decisions.
Open manual →Occupation manualA regulated, proof-first AI workflow for advisor meeting preparation, privacy-safe input control, neutral question queues, and supervised review—without individualized advice or recommendations.
Open manual →Occupation manualA proof-first AI workflow for property intake, emergencies, fair housing, work orders, inspections, vendors, resident messages, budgets, and incidents.
Open manual →Occupation manualA source-controlled AI workflow for appraisal intake, comparable research preparation, report QA, confidentiality, bias checks, UAD 3.6 readiness, and licensed review.
Open manual →Occupation manualA proof-first retail AI workflow for shift briefs, source checks, exception lists, data boundaries, and human-reviewed handoffs.
Open manual →Occupation manualA source-bound AI workflow for technical discovery, capability gaps, demo outlines, review owners, and safer pre-sales handoffs.
Open manual →Occupation manualA review-only AI workflow for sales managers: prepare source-bound forecast and coaching packets while keeping CRM access, scores, decisions, and communications human-controlled.
Open manual →Occupation manualA proof-first workflow for useful answers, source coverage, crawler controls, canonical and sitemap parity, structured data, submissions, citations, and refreshes.
Open manual →Occupation manualA proof-first AI coding workflow for requirements, change plans, tests, security, accessibility, review, release, rollback, and monitoring.
Open manual →Occupation manualA proof-first AI workflow for approved test design, exact build identity, execution receipts, defects, traceability, flaky tests, and release-review handoffs.
Open manual →Occupation manualUse AI to prepare website change packets, QA checks, rollback notes, and incident handoffs without handing it production control or secrets.
Open manual →Occupation manualA source-bound AI workflow for wholesale and manufacturing sales account briefs, unknown logs, review questions, and human approval handoffs.
Open manual →Define one bounded job, people/rights/money/property affected, approved systems, and the qualified reviewer.
Separate law, rule, program, employer policy, governing document, qualified record, participant statement, and AI summary.
Build an evidence table, checklist, draft, QA report, exception queue, or exact handoff without claiming execution.
Keep regulated judgment, rights, safety, money, messages, accounts, production, and public actions with authorized humans.
If the profession is not on the completed shelf, use the route card to define the job and keep the request on the research board until a real evidence-backed manual exists.