Quick snapshot of your current underwriting flow

• Tell me about your current submission intake channels and volumes, which channels deliver the most commercial submissions? Options: Email attachments, Broker portal uploads, Direct portal/API, File feeds (SFTP, batch), Paper or scanned documents, Other
• In a typical month, how many broker submissions does your commercial lines team receive? Options: Under 100, 100–499, 500–1,999, 2,000–9,999, 10,000+
• Walk me through the sequence an underwriter follows from receiving a broker PDF to producing a first price indication, including who touches the file and which systems get updated.
• Who on your team has final sign-off for coverage and pricing decisions on typical commercial submissions? Options: Chief Underwriting Officer, Head of Commercial Lines, Line Underwriting Manager, Individual underwriters by authority tier, Shared committee
• On average, how many business days elapse from initial receipt to the first underwriting decision for a standard commercial account? Options: Same day, 1–3 business days, 4–7 business days, 8–15 business days, More than 15 business days

Where manual work erodes time and accuracy

• Which manual step, if automated, would free the most underwriting capacity in your commercial lines? Options: Manual data entry from broker PDFs, Manual retrieval of loss runs, Manual property characteristic lookups, Manual re-keying into rating engine, Routing work to the right underwriter
• Describe the typical errors from manual entry that lead to pricing mistakes, rescopes, or delayed binding.
• Walk me through a recent submission where missing or incorrect broker information materially changed the outcome, and tell me who absorbed the remediation cost.
• Roughly how often do underwriters escalate submission exceptions to a manager because of incomplete data? Options: Almost every day, Several times a week, Weekly, Monthly, Rarely
• If integration to your rating engine or policy administration cannot be completed during pilot, would that prevent you from running the pilot or only change its scope? Options: Prevent pilot entirely, Change pilot scope only, Run with manual workarounds, Undecided

Where underwriter adoption breaks down

• When a new intake workflow is introduced, which reason do your most experienced underwriters most often give for not adopting it? Options: Loss of control over judgment, Extra clicks or friction, Poor data quality from automation, Lack of trust in AI outputs, Insufficient training/time
• Tell me about the last time you tried to automate intake or worksheet population, what worked, and what caused the rollout to stall.
• Which training format historically produces the fastest underwriter proficiency: live workshops, one-on-one shadowing, short micro-lessons, or embedded help in the tool? Options: Live workshops, One-on-one shadowing, Micro-lessons (5–10 min), Embedded in-app help, Blended approach
• List the top three metrics leadership uses today to judge whether an intake or worksheet change is successful.
• If a pilot shows a 50% reduction in manual entry time but an initial 10% drop in hit rate, would you move forward to production, pause for adjustment, or stop the evaluation? Options: Proceed to production, Pause and optimize, Stop evaluation, Need more context

Imagine a faster, clearer underwriting day

• Describe the single change to an underwriter's daily workflow that would make you convinced modernization is worth the effort.
• In a scenario where submissions arrive pre-populated and scored, how many additional accounts could an average underwriter handle per month? Options: Fewer than 5, 5–20, 21–50, 51–100, 100+
• Identify the decision points you believe must remain human even with AI assistance, and why. Options: Coverage decisions, Authority exceptions, Complex risk judgment, Pricing overrides, Binding decisions
• To what extent would real-time pipeline visibility change how you reallocate capacity across product lines? Options: Radically, Significantly, Moderately, Slightly, Not at all
• Assuming the pilot demonstrates the projected throughput gains, how quickly could you scale the solution to another line of business? Options: Within 30 days, Within 2–3 months, Within 4–6 months, Longer than 6 months, Not sure

Integration and data readiness, the hard gating questions

• Name the single integration or data capability that, if unavailable, would stop the project before it starts.
• Who owns API access and integration support for your policy administration, rating engine, and document storage systems? Options: Internal integration team/APIs team, Platform engineering, Third-party managed services, Line IT manager, Other
• How many dedicated technical FTEs can your organization assign to implementation and testing during a 6–8 week pilot? Options: None, 1–2, 3–5, 6–10, More than 10
• Does your data retention and privacy policy permit sharing broker submission PDFs with a hosted AI extraction service under a standard data processing agreement? Options: Yes, No, Needs legal review
• Suppose API access cannot be provided within your target 8-week timeline, would you prefer to postpone, reduce pilot scope, or run with manual workarounds? Options: Postpone pilot, Reduce scope, Run with manual workarounds, Cancel evaluation

What alternatives are you actively weighing

• Point to the external vendors, incumbent platforms, or internal teams you are considering as alternatives, and note which is currently the frontrunner.
• For each option you listed, what single advantage would make your organization stay with that option instead of selecting an outside platform?
• Rank the reasons that would keep you on your current approach, select all that apply. Options: Lower immediate cost, Existing integrations in place, Internal control/preference, Perceived lower change risk, Vendor relationship/trust
• Has anyone proposed building this capability internally instead of buying, and if so, who would lead that effort? Options: Yes, IT/Engineering, Yes, Insurtech/Product, No one has proposed it, Other
• What would have to be true about the incumbent or internal build to make you cancel this evaluation and stick with the current approach today?

Pilot success criteria and go/no-go moments

• Define the pilot outcome that would lead your team to sign a production contract within 30 days of acceptance.
• Provide the acceptance thresholds you expect for extraction accuracy, routing correctness, and end-to-end decision time improvement (use percentages or time windows).
• How many underwriters must participate in the pilot for the result to be statistically and operationally credible? Options: 26–50, 1–3, 4–10, 11–25, 50+
• Identify the person or role who will sign off on pilot acceptance and describe the authority they hold to commit budget or schedule.
• In the event the pilot misses a key acceptance metric by 10%, would you require a re-run, accept with concessions, or stop the evaluation? Options: Require a re-run, Accept with concessions and mitigation plan, Stop evaluation

Timeline, purchasing cadence, and blockers

• Point to the procurement or legal hurdle most likely to delay contracting beyond 90 days.
• Typically, how many business days does your legal team take to complete a standard vendor agreement review? Options: Under 10 days, 10–20 days, 21–40 days, 41–90 days, Longer than 90 days
• Name the budget holder who approves subscription and integration costs for adding another production line. Options: CFO/Finance, Chief Underwriting Officer, Head of Commercial Lines, IT Budget Owner, Shared approval
• To move from pilot to contract signing, what is the single deliverable or decision you need from the seller?
• Does your finance team have five quarters of P&L history and topline metrics available to model uplift if the board requests an ROI estimate? Options: Yes, No, We can assemble on request

Final readiness and next-step commitments

• Are there any organizational changes, ownership gaps, or upcoming initiatives that would prevent a rollout even if the pilot succeeds?
• Specify the stakeholders who must be engaged before production go-live and the primary responsibility each will own (integration, training, operations, legal). Options: Underwriting leadership, IT/integration team, Operations/processing, Legal/compliance, Finance, Training/Change Management
• Can you provide two weeks of historical broker submissions for model tuning and validation within 10 business days if requested? Options: Yes, No, Need approval from data owner
• Are you ready to schedule a technical kickoff within 14 days if pilot scope and commercial terms meet your targets? Options: Yes, Maybe with conditions, No

• Solution Experience Session

• Confirm the current state and its cost to your team
• Customer confirms the demonstrated workflow removes the manual rekeying and rework described in Discovery.
• Seller to run batch extraction on five representative submissions provided by the customer and deliver an accuracy comparison report before the follow-up session.
• Customer confirms AI extraction accuracy on the sample submission is within an acceptable threshold or defines the acceptable threshold.
• Run your submission end-to-end through intake and worksheet
• Seller to provide a proposed pilot plan with success metrics, timeline, and resource requirements.
• Customer to provide five representative broker submissions in original format and identify the pilot underwriter group.
• Agreement on pilot success metrics, representative underwriter group, and the list of integration owners required to start a pilot.
• Show rating and reporting behavior using your scenario
• Customer to identify integration owners and share API documentation or access details for policy administration and rating engine endpoints.
• Identify remaining technical or organizational objections that would block a pilot decision.
• Confirm integration and exception handling for your systems
• Validate that the demonstrated future state matches your need
• Customer to confirm the acceptance thresholds for AI extraction accuracy and acceptable cycle time improvements for pilot approval.
• Agree next evidence and pilot metrics

• Solution Experience Session
• Solution Experience Deck
• Solution Brief

• meeting
• slides
• document

• Automated Submission Intake and Triage
• AI-Powered Data Extraction from Broker Documents
• Pre-Populate Underwriting Worksheets
• Third-Party Data Enrichment Connectors
• Configurable Underwriting Rules and Authority
• Automated Risk Scoring and Prioritization
• Underwriter Routing and Capacity-Based Assignment
• Quote Generation and Bind Workflows
• Policy Administration System Integration
• Rating Engine Integration
• Document Management and Storage Integration
• Management Dashboards for Pipeline and Productivity
• Migration of Historical Submission and Pipeline Data
• Underwriter Workspace Configuration and Training

Automated Submission Intake and Triage

• Do you receive submissions primarily by email, portal upload, or broker FTP, and which formats (PDF, XML, images) are most common? Options: Email (PDF attachments), Portal uploads (PDF/ZIP), Broker FTP/SFTP, Mixed formats including images/photos
• Which broker submission types account for the majority of volume (e.g., ACORD commercial package, Certificate of Insurance, loss runs)? Options: ACORD commercial package, Certificates of Insurance (COI), Loss runs, Non-ACORD bespoke submissions
• How many broker submissions does your underwriting organization average per week across the targeted lines of business? Options: Less than 250, 250-1,000, 1,000-5,000, More than 5,000
• Describe the current triage rules you use to prioritize and route new submissions (e.g., by line, limit, broker tier, territory).
• Who will be the nominated owner in your organization for intake exceptions and manual triage decisions during pilot?
• Provide examples of metadata or tags you need captured on intake (e.g., broker name, submission type, requested effective date, indicated limit).

AI-Powered Data Extraction from Broker Documents

• Which specific fields must be extracted from broker PDFs and loss runs for automated prefill (for example: policy effective date, payroll, building limit, prior carrier, total payroll)?
• How much variability do your broker documents exhibit (standard ACORD forms vs. scanned bespoke submissions) and what percentage are scanned images versus native PDFs? Options: Mostly native ACORD PDFs, Mixed native and scanned, Mostly scanned images or photos, Unknown
• Do you require extraction of structured lists such as schedules of locations, vehicle lists, or named insured subsidiaries from attachments? Options: Yes, locations schedule, Yes, vehicle schedules, Yes, subsidiary lists, No structured lists needed
• Specify the target field-level extraction accuracy you require for go/no-go decisions on submission triage (give a percentage for key fields such as limits, effective date, and payroll).
• What acceptance criteria will confirm that AI extraction is acceptable for pilot (for example: X% accuracy on core fields, Y% coverage of submissions)? Options: Accuracy >= 90% on core fields, Coverage >= 85% of incoming submissions, Both accuracy and coverage targets must be met
• Provide a prioritized list of up to 10 fields you want extraction to focus on first for pre-populating underwriting tasks.

Pre-Populate Underwriting Worksheets

• Which underwriting worksheet templates do you use today and which worksheet fields must be pre-populated from extracted submission data (for example: exposure bases, limits, deductibles, retro dates)?
• How many distinct product lines or worksheet variants need separate field mappings (for example: commercial property, general liability, cyber)? Options: 1, 2-3, 4-6, 7+
• List any computed fields or business rules that must be applied after pre-population (for example: total insured value calculations, payroll band assignments).
• Who in your underwriting team will own validation of pre-populated worksheets and correction of extracted values during the pilot?
• Indicate how often worksheet templates change in your shop (monthly, quarterly, annually) and whether you need a UI to edit templates without engineering support. Options: Monthly, Quarterly, Annually, Rarely
• Provide 2-3 sample underwriting worksheet PDFs or screenshots that show non-standard fields we must support. Options: I will upload samples, I need assistance to collect samples

Third-Party Data Enrichment Connectors

• Which external data sources do you rely on during underwriting (for example: claims loss history API, property characteristic feeds, credit/financial feeds)?
• Do you require enrichment of loss runs via a loss-history normalization process or direct display of carrier-provided loss run documents? Options: Normalized loss-history summary, Direct display of original loss run documents, Both
• Identify the integration method available for each connector: API key, SFTP batch, or vendor-hosted portal. Options: API key/REST, SFTP/flat-file, Vendor-hosted portal, Not available
• How frequently must third-party data refresh for underwriting decisions (real-time, hourly, daily, on-demand lookup)? Options: Real-time / on-demand, Hourly, Daily, Batch weekly
• Describe any regulatory or data residency constraints for external enrichments (for example: EU data residency, sensitive personal information restrictions).
• List the top three data elements from third-party feeds that will directly affect score or routing (for example: prior loss frequency, building year, credit rating).

Configurable Underwriting Rules and Authority

• What underwriting authority tiers exist in your organization (for example: junior underwriter up to $250k, senior up to $1M, delegated bind authority)?
• How should appetite and declination rules be expressed (rule-builder UI, coded rules, or spreadsheet import)? Options: Rule-builder UI, Coded/business rule config, Spreadsheet import
• Which specific risk characteristics drive automatic declination or referral (for example: payroll > X, prior carrier bankruptcies, specific NAIC/industry codes)?
• Who will own rule changes in production and what is the approval workflow for editing authority thresholds?
• Identify the number of authority tiers and any dollar thresholds you want enforced in routing and bind controls.
• Describe how exceptions are documented today (email thread, spreadsheet, custom ticket) and whether you need exception audit trails retained in the platform. Options: Email/thread, Spreadsheets, Ticketing system, Need audit trail in platform

Automated Risk Scoring and Prioritization

• Which scoring inputs must be included in the prioritized queue (for example: indicated premium, loss severity, broker tier, renewal vs new business)?
• How should risk priority be surfaced to underwriters (sorted queue, alert badges, SLA flags)? Options: Sorted queue, Alert badges, SLA deadline flags, All of the above
• Specify any mathematical model or threshold used today for prioritization (for example: score > 75 routed to senior underwriter).
• Who will validate the scoring model outputs and what sample size do you require for statistical validation during pilot?
• List the KPIs that prioritization must improve or report on (for example: cycle time reduction, hit ratio improvement, percent of submissions reviewed within SLA).
• Describe any seasonal or portfolio-level weighting that should influence scoring (for example: higher weight to property risks in hurricane season).

Underwriter Routing and Capacity-Based Assignment

• Do you currently manage underwriter capacity with numeric limits (open submissions per underwriter) or by workload hours, and what are those thresholds? Options: Open submissions limit, Workload hours estimate, Hybrid
• Which routing rules do you need supported (by appetite, geography, capacity, specialty, or broker relationship)? Options: Appetite, Geography, Capacity, Broker relationship, Specialty
• How many routing exception scenarios require managerial override and who should be notified when an override occurs?
• Indicate whether you need automatic rebalancing of assignments during absence or planned leave periods for underwriters. Options: Yes, automatic rebalancing, No, manual reassignments only, Prefer configurable option
• Provide the typical SLA for an underwriter to respond to a newly assigned submission (for example: 24 hours, 48 hours). Options: 4 hours, 24 hours, 48 hours, Custom
• Who will own capacity rule configuration and what governance do you require for changing routing policies?

Quote Generation and Bind Workflows

• Which parts of quote creation need automation (rate calculation, policy document assembly, quote letter generation, carrier submission)? Options: Rate calculation, Policy document assembly, Quote letter generation, Carrier submission
• Do you require templated quote letters and bind documents with variable fields drawn from pre-populated worksheets? Options: Yes, templated letters, No templates needed, Partial templating required
• Describe any manual approval gates required before a quote can be issued or bound (underwriter sign-off, manager approval, compliance).
• Indicate whether certificate issuance and COI automation should be part of the bind workflow. Options: Yes, COI automation, No, manual COI issuance, Partial automation
• List post-bind operational steps that must be triggered (policy admin create, billing handoff, document archival).
• Who should be notified automatically upon bind (producer, broker, operations team) and by what channel (email, API callback)? Options: Broker email, Producer email, Operations notification, API callback

Policy Administration System Integration

• Which policy administration system will you integrate with and what integration method is available (REST API, SOAP, file drop)? Options: REST API, SOAP, File drop / SFTP, No direct integration available
• Specify the key API endpoints required for go-live (policy create, endorsements, cancellations, policy retrieval).
• What data mapping complexities exist between worksheet fields and policy administration fields (for example: nested coverages, multi-location schedules)?
• What acceptance criteria will demonstrate successful end-to-end integration with the policy admin system (for example: X policies created, Y endorsements processed without error)? Options: 100 policies created in test without errors, All endorsements processed with zero critical errors, API end-to-end response time < 3s
• Who is the integration owner on your side (API team contact) and who will provide test credentials and sandbox access?
• State any transformation or enrichment that must occur before payloads are sent to the policy admin (for example: normalize location addresses, map rating codes).

Rating Engine Integration

• Which rating engine interface is available for your team (real-time API, batch file import, or hosted calculation service)? Options: Real-time API, Batch file import, Hosted calculation service, Not yet determined
• How many distinct rating schemas or rule sets must the integration support (for example: GL, Property, Auto rating logic)? Options: 1, 2-3, 4-6, 7+
• List the key rating inputs we must supply from the worksheet (for example: exposure bases, class codes, limits, deductible).
• Indicate the acceptable latency for a rating call during quote flow (for example: <1 second, <3 seconds, asynchronous batch). Options: <1 second, <3 seconds, Asynchronous batch
• Who will own rate table updates and how are rating changes communicated to the integration team?
• Specify any calculation audit requirements (for example: store inputs/outputs for each rated quote for X days).

• current_state
• success_criteria
• decision_readiness
• gaps
• stakeholders
• desired_state
• gaps
• current_state
• desired_state
• success_criteria
• stakeholders
• decision_readiness
• current_state
• stakeholders
• decision_readiness
• success_criteria
• gaps
• decision_readiness
• decision_readiness
• decision_readiness
• decision_readiness

• Subscription Agreement
• Order Form (Pricing & Billing)
• Master Services Agreement (MSA)
• Statement of Work (SOW) — Implementation & Pilot
• Pilot Acceptance Certificate
• Production Rollout Responsibilities
• Security & Data Processing Addendum (DPA) — SOC2 / Data Residency
• Change Order Agreement

• Go-live Health Check (weeks 1-4)
• First Outcomes Review (weeks 4-10)
• Acceptance Decision and Remediation (around day 90)
• Operational Monthly Review (ongoing monthly)
• Quarterly Success Review (quarterly executive and program owners)

• Agree the prioritized set of enhancement requests to be actioned in the next quarter.
• Restate acceptance criteria and targets
• Produce a documented acceptance decision against the acceptance criteria and numeric targets recorded in Pilot Evaluation, with a named signatory where required.
• For any failed acceptance criteria, agree remediation plans, verification tests, and hard completion dates.
• Confirm production readiness or clearly documented conditional acceptance with defined rollback or mitigation steps.
• Publish the formal acceptance decision document and distribute to all stakeholders.
• For each failed criterion, publish a remediation plan with verification criteria and target completion dates.
• Schedule a verification session on each remediation completion date to confirm the criterion is met.
• Performance trend review
• Confirm whether cycle time and underwriter throughput are improving toward the targets recorded in Pilot Evaluation.
• Reduce the number of aged high-severity incidents and assign remediation actions with deadlines.
• Re-confirm success criteria and owners
• Assign owners and due dates for all incidents older than 7 days and publish the burn-down plan.
• Produce a prioritized enhancement backlog with business impact and estimated delivery effort.
• Schedule and deliver a targeted training session for underwriters with below-threshold throughput metrics.
• Executive metrics summary
• Confirm adoption and business outcome metrics are tracking to the targets recorded in Pilot Evaluation or surface an escalation plan if not.
• Ensure no legacy system fallback behavior persists and that archival and decommission tasks are closed.
• Agree the operational priorities and verification checkpoints for the next quarter.
• Deliver the consolidated quarterly metrics report with trendlines and variance explanations to stakeholders.
• Close any remaining legacy fallback or shadow-work incidents by the agreed date.
• Publish the next quarter's operational commitments and verification checkpoints.
• Confirm the platform is functionally live for all core underwriting intake and routing workflows.
• Confirm the incumbent submission process is on a decommission or read-only trajectory and no shadow work persists.
• Assign owners and deadline dates for all high-severity issues identified during go-live validation.
• Produce a short go-live validation checklist with pass/fail status for each integration and core workflow.
• Publish the incumbent wind-down status and any outstanding dependencies toward read-only or decommission.
• Create an owners-and-dates list for all high-severity issues surfaced in the session.
• Present first-measured KPI results
• Determine whether AI extraction accuracy (%) and average manual data-entry time per submission (minutes) meet or are trending to the targets recorded in Pilot Evaluation.
• Identify top 2 root causes for any KPI shortfalls and agree concrete remediation steps and dates.
• Confirm readiness status and timeline for the acceptance decision meeting.
• Run a field-level extraction error analysis for the top 5 failed fields and produce remediation recommendations.
• Deliver an integration error-rate report covering the previous 30 days with suggested fixes and timelines.
• Schedule a targeted training module for underwriters identified with low proficiency within 7 days.
• Platform performance and integration uptime
• Operational incidents and ticket burn-down
• Deployment and migration validation
• Root-cause diagnosis for KPI gaps
• Present outcome data per acceptance criterion
• Early adoption signals and usage patterns
• Long-running issues and enhancement status
• Document formal acceptance decision
• Enhancement request backlog and prioritization
• Integration stability review
• Incumbent system wind-down confirmation
• Adoption and training gaps
• Confirm legacy fallback and data archival status
• Training and adoption follow-up
• If failed, agree remediation plan and verification steps
• Agree next quarter operational commitments
• Open issues and blockers
• Confirm any outstanding incumbent wind-down items
• Agree corrective actions and timeline to acceptance
• Agree next month's operational targets
• Agree immediate remediation actions