Callright — The Community-Bank Call Report & NCUA 5300 Reporting Engine

01Executive summary

Every one of the ~8,800 U.S. banks and credit unions must file a full regulatory financial report every quarter — the FFIEC Call Report for banks and NCUA Form 5300 for credit unions — under statutory mandate, with civil money penalties for late or false filings Verified. The report is a multi-thousand-data-item, cross-validated document that must reconcile to the GL and pass hundreds of published FFIEC edit checks. At small institutions the person who "owns" it is often one overworked controller or cashier, and community-bank finance talent is in structural short supply Verified.

Today's incumbents (Wolters Kluwer, Fiserv, FedReporter/SmartCall, DBI EasyCall, Regnology, Nasdaq) sell software the bank's own staff must operate Verified. Callright sells the outcome: the bank uploads its trial balance and core-system extracts, and receives a finished, edit-clean, variance-explained Call Report package plus the CDR-ready file — reviewed and signed off by a licensed preparer — for its authorized officer to certify and submit. AI performs extraction, GL-to-schedule mapping, edit-check resolution, peer-benchmark anomaly detection, and variance-narrative drafting; a former bank accountant reviews only the exceptions.

02Thesis

Regulatory financial reporting for depository institutions is a recurring, mandatory, high-consequence, highly structured workflow — the ideal substrate for an AI-native service. The task decomposes cleanly into extraction, mapping, deterministic validation, and a narrow band of judgment (classification edge cases, material-variance explanations, new-instruction interpretation). Frontier models make the automatable 90% cheaper and faster every quarter; a licensed reviewer owns the judgment 10%. Because the institution's officer certifies and submits, Callright bears preparation accountability without stepping into attest or legal territory.

We win not by being "AI for call reports" but by selling a filed, examiner-defensible outcome with an amendment-free standard — something no self-service software vendor sells, and something CPA/advisory firms sell only at bespoke, labor-heavy prices. The moat compounds: a growing library of GL-to-schedule mappings across core systems and chart-of-accounts variants, a quarterly-maintained FFIEC edit-check and instruction engine, and an audit trail examiners trust.

03Discovery rationale

This run's search terrain spanned banking/fintech operations, tax/audit compliance, healthcare admin, insurance, and government paperwork. Three observations drove selection:

1. The 2025–2026 environment is unusually deregulatory. Two attractive-looking candidates — SEC inline-XBRL/EDGAR filing and the FinCEN Residential Real Estate report — were weakened or killed during research: the SEC has proposed optional semiannual reporting (reducing per-filing volume) Verified, and the FinCEN Residential Real Estate Rule was vacated by a federal court in 2026 Verified. Lesson: anchor on a statutorily permanent mandate, not a fragile new rule.
2. The Call Report / 5300 mandate is about as permanent as U.S. financial regulation gets. It predates every current agency leader and survives every deregulatory cycle; the 2025 RFI seeks to streamline, not eliminate, it Verified.
3. The true demand driver is labor, not a single rule. Community-bank compliance already consumes 11–15.5% of personnel expense at small institutions, and the talent pool is structurally shrinking Verified — a durable tailwind independent of regulatory direction.

04Candidate comparison

Five candidates were generated and scored (1–5, higher better) on the dimensions most predictive of a foolproof AI-native service. The Call Report engine wins on mandate durability and demand evidence while sharing the same structured/regulated shape.

*FinCEN RRE disqualified: rule vacated by federal court in 2026 (fatal disqualifier — no live mandate). SEC XBRL de-prioritized due to the semiannual-reporting proposal reducing per-filing volume. PCI DSS overlaps a prior blueprint (SOC 2 readiness). SERFF is a viable second product but has thinner public demand evidence than the Call Report labor crisis.

05CODE validation

06Rubric scorecard — the six gates

Composite: 29/30. Pricing is per-unit (per accepted filing), never hourly.

07Target buyer

08Jobs-to-be-Done

• Functional: "File an accurate, edit-clean Call Report on time every quarter so we avoid a CMP, an amendment, or an examiner finding."
• Functional: "Explain every material quarter-over-quarter variance before the examiner asks."
• Emotional: "Stop the panic when the one person who knows the Call Report is out, quits, or retires."
• Social: "Look competent and in control in front of my board and my examiner."
• Economic: "Free 40–80 skilled hours per quarter for work that actually grows the bank."

09The painful problem

The Call Report is a multi-thousand-data-item quarterly filing spanning ~30 schedules (balance sheet RC, income RI, loans RC-C, past-due RC-N, deposits RC-E, regulatory capital RC-R, and more). It must reconcile to the general ledger to the dollar and pass hundreds of published FFIEC validity and quality edit checks before the CDR will accept it Verified. Instructions run hundreds of pages and are revised nearly every quarter Verified.

At a small bank this lands on one or two people. Errors trigger amendments, examiner Matters Requiring Attention, and — for late or knowingly false filings — tiered civil money penalties Verified. When that person leaves (and community-bank turnover is worsening Verified), the institution faces an existential quarterly deadline with no bench. Software doesn't solve this: it still requires a trained operator to map accounts, resolve edits, and defend variances.

10The outcome we sell

Not a tool. A filed-ready, examiner-defensible Call Report, every quarter, on time.

Deliverable per quarter: (1) the CDR-ready data file (validated against current FFIEC edit specs), (2) a human-readable review package showing every schedule with prior-period comparison and a written explanation of each material variance, (3) an exceptions/assumptions log listing every judgment call for officer confirmation, and (4) an audit trail linking each reported figure back to its GL/source. The bank's authorized officer reviews, certifies, and submits. Callright's standard: zero edit-check failures at submission and zero examiner-required amendments, or we remediate at no charge.

11First one-feature MVP wedge

12Evidence summary

13Claim table (Verified / Inferred / Unverified)

14Source-claim matrix

15Market & demand evidence

The addressable filer base is fixed and mandatory: ~4,462 FDIC-insured banks plus ~4,331 federally insured credit unions Verified, each filing four core quarterly reports plus supplementals. That is ~35,000 core filings/year with no substitution: you cannot choose not to file. Demand is corroborated by (a) an entrenched paid software market (six-plus certified vendors) Verified; (b) CPA/advisory firms selling outsourced completion Verified; and (c) a $19B RegTech market growing ~16% CAGR Verified. Overlaid on this is a labor supply shock (11–15.5% of personnel cost on compliance; structural talent shortage) that pushes institutions from "buy software, staff it ourselves" toward "buy the outcome" Verified.

16Active buyer conversations

Buyer pain is publicly visible in: the OCC/Fed/FDIC RFI comment process where banks are actively documenting Call Report burden [S1]; state bankers-association forums and ICBA advocacy on regulatory burden; core-processor and reporting-software user communities troubleshooting edit-check failures each quarter; and CPA-firm marketing pages that exist precisely because banks ask them to "complete the Call Report" [S13]. Job postings for "regulatory reporting analyst / Call Report specialist" and their scarcity are the clearest labor-side demand signal Inferred.

17Competitive landscape

18Competitor & budget validation

Budget already exists and is redirectable. Banks currently pay: (1) annual software license fees to certified vendors Verified; (2) fully loaded salary + benefits for a Call Report preparer (a scarce, rising cost) Verified; and (3) advisory/CPA fees when they outsource completion or need help after a finding Verified. Callright's per-filing price sits below the fully loaded cost of the internal FTE and competes on productization/price against senior-hour CPA work. "No competitors" is not the thesis — the thesis is a large, entrenched, budgeted category served only by tools and bespoke labor, with an unoccupied productized-outcome slot.

19Pricing evidence & proposed pricing

Public per-filing pricing is scarce (vendors negotiate) Inferred, but the anchors are clear: small institutions spend low-five-figures/year on reporting software and materially more on the internal FTE. Callright prices per accepted filing, aligned to outcome, never hourly:

Optional amendment-free service standard: if an examiner requires an amendment to a Callright-prepared schedule, we remediate free and credit the next filing. This monetizes accountability without contingency/UPL risk.

20Regulatory & compliance considerations

The Call Report and Form 5300 are agency-mandated financial reports. Preparation by a third party is well-established and permitted; the institution's authorized officer certifies and submits Verified. Key considerations: (1) Bank data is highly sensitive — GLBA/Reg P, examiner expectations for vendor risk management (the bank must be able to include Callright in its third-party risk program); (2) examiner defensibility — every figure must trace to source with an audit trail; (3) instruction currency — reporting must reflect the current-quarter FFIEC instruction set, revised frequently Verified; (4) no attest — Callright prepares, it does not audit or issue an opinion, avoiding CPA-attest and independence entanglements.

21Licensing boundary

22AI-native advantage

"AI-native" here means AI changes the economics and speed of production, not that the customer touches a chatbot. The engine ingests messy trial balances and core-system exports (Fiserv, Jack Henry, FIS variants), maps them to schedules using a learned + rules-based mapping library, runs the deterministic FFIEC edit-check suite, benchmarks each line against the institution's own history and UBPR-style peer norms to surface anomalies, and drafts material-variance narratives. What must never be fully automated: final classification judgment on edge cases, interpretation of new instructions, and the sign-off. Failure risks (mis-mapping, stale instruction set, silent GL change) are caught by deterministic checks + reviewer, not left to the model.

23Internal AI engine architecture

24AI-vs-human operations pipeline

25Dynasty translation layer

26Anti-duplication analysis

What exists: customer-operated reporting software (WK, Fiserv, FedReporter, DBI, Regnology) and bespoke CPA/advisory outsourcing. Why this is not a copy: it is neither a tool nor senior-hour labor — it is a productized, per-filing, AI-produced outcome with an amendment-free standard. Narrow wedge: FFIEC 051 small banks with a departed preparer. Under-served segment: $100M–$1.5B institutions too small for enterprise software teams and too price-sensitive for Big-4-style advisory. Unsolved pain: the labor/knowledge single-point-of-failure that software cannot remove. Differentiation moat: the cross-core mapping library, the quarterly-maintained edit-check/instruction engine, examiner-grade audit trails, and accountability that vendors explicitly disclaim. Checked against the 110 prior blueprints (incl. HMDA LAR integrity, ASC 740/842, ERISA 5500): none address the quarterly Call Report / 5300 core financial filing — this is fresh.

27Anti-commoditization analysis

If future general models make trial-balance reading trivial, Callright still wins because the product is accountability + the compliance operating system, not the extraction. A bank cannot let a general model file its Call Report unsupervised: it needs the examiner-defensible audit trail, the current-quarter edit-check engine, the peer-anomaly benchmark, the reviewer sign-off, and someone to stand behind an amendment. The proprietary assets — mapping library across cores, correction history, examiner-outcome data, and the instruction-change pipeline — deepen with volume and are not reproduced by a smarter base model. Commoditization of the easy 90% increases our margin; it does not remove the reason to buy.

28Service delivery workflow

1. Onboarding (once): capture charter, form type, core system, chart of accounts, prior 2 filings; build baseline mapping.
2. Quarter open: bank uploads trial balance + core extracts to secure portal (or scheduled SFTP/API pull).
3. Prep: engine normalizes, maps, computes, and runs edit checks; flags exceptions.
4. Review: reviewer clears exceptions, approves variance narratives, finalizes numbers.
5. Deliver: package + CDR file + exceptions log + audit trail sent ≥5 business days before deadline.
6. Certify: bank officer reviews, certifies, submits; Callright provides submission-support if edits arise at CDR.
7. Post-file: archive; capture any examiner feedback; feed learning loop; schedule next quarter.

29Operations as product

The operation is the product. Variance elimination is enforced by: SOPs per form type; structured intake checklists (no prep starts without a complete input set); required-evidence lists per schedule; automated completeness checks; an exceptions queue with reviewer assignment logic; confidence scoring on each mapping; full audit trails and version control; gold-standard reference filings; quarterly red-team review against new instructions; customer-ready package templates; and root-cause analysis + postmortem for any edit failure, amendment, or missed deadline. Every defect becomes a new rule, check, or template.

30No-holes quality engine

• Deterministic gate: the filing cannot be delivered until it passes the full published validity + quality edit-check suite with zero failures.
• Reconciliation gate: every schedule ties to the GL and to prior period; unexplained deltas block release.
• Peer-anomaly gate: lines that deviate from the institution's history or peer norms are surfaced for explicit confirmation.
• Human gate: a reviewer signs off on classifications, material variances, and any new-instruction interpretation.
• Exceptions transparency: every judgment call is logged and confirmed by the bank before certification.

31What the human expert actually does

Target reviewer time: ~170 min/filing at launch → ~65 min at day 90. Documentation/audit trail auto-generated for every task.

32Minimum viable offer

"We prepare your FFIEC 051 Call Report every quarter — edit-clean, variance-explained, and delivered five days early — for a fixed price. If an examiner ever requires an amendment to a schedule we prepared, we fix it free." One form type, one ICP, one measurable outcome (0 edit failures, 0 amendments, on time).

33Fulfillment process (first 3 customers, semi-manual)

1. Analyst collects inputs; builds mapping in a working spreadsheet + LLM-assisted extraction.
2. Run edit checks via the vendor CDR validation / a scripted checklist against published edit specs.
3. Reviewer (fractional CPA with bank experience) verifies numbers and variances.
4. Deliver package + CDR file; support the officer through certification.
5. Debrief every filing; codify each mapping and exception into SOPs and the nascent engine.

Day-one tools only. Automate mapping + edit-checks after ~5 filings reveal stable patterns.

34Tools & systems

Secure intake portal + SFTP; document parsing/OCR; LLM API (model-agnostic) for extraction, mapping, and narrative drafting; a deterministic edit-check engine built to the published FFIEC/NCUA specs; a review workbench with exceptions queue and audit trail; CDR-compatible file exporter; CRM + engagement tracker; encrypted storage with access controls and a vendor-risk-management packet for banks' exam files. Buy/assemble before building bespoke software; the differentiator is the mapping/edit knowledge base, not the UI.

35Human-in-the-loop quality control

Two independent controls guarantee quality: (1) a machine gate — deterministic edit checks + reconciliation that must pass before a human sees the file; and (2) a human gate — a reviewer who signs off on judgment items and releases the exceptions log. High-risk schedules (RC-R capital, RC-N past-due, RC-C loan codes) get mandatory second-look. Confidence scores route low-confidence mappings to senior review. The bank's own certification is the final control.

36Nonlinear scaling & unit economics

Automation share: ~55% at launch → ~80% at day 90 → ~90% at year 1. Throughput: ~2–3 filings/reviewer-day at launch → ~8–10 at day 90. Cycle time: 3–5 business days. CAC payback target < 2 filings given quarterly recurrence. Conversion assumptions: lead-magnet→consult 8–12%; waitlist→pilot 30–40%; pilot→paid 50–65%; annual retention 85–90% (quarterly habit + switching cost of re-teaching mappings). All figures modeled — Inferred/Unverified — to be replaced with pilot actuals.

37Distribution proof table

38Sales & outreach plan

Three-layer motion: (1) expert-led content teaching the buyer why the Call Report is their biggest single-point-of-failure and what edit errors cost; (2) warm GTM to webinar/diagnostic/calculator users and association contacts with a scoped pilot offer; (3) targeted outbound to CFOs/controllers at 051-filing banks, leading with a personalized "reporting continuity" diagnosis, not a demo ask. Partnerships with bankers' banks and bank CPA firms are the highest-leverage channel because they aggregate the exact ICP and already carry trust.

39Founder-led / expert-led content plan

Publish from a named ex-bank-CFO or ex-examiner voice on: the true cost of a Call Report amendment; the RC-R capital schedule mistakes examiners flag most; how instruction changes each quarter create silent errors; building reporting continuity when your one preparer leaves; what "quality edits" vs "validity edits" really mean; and quarter-close checklists. Repurpose top organic posts as paid-ad creative later.

40First 30 days of content

10 educational posts: (1) The $X cost of a Call Report amendment; (2) Why RC-R trips up small banks; (3) Validity vs quality edits, explained; (4) The single-point-of-failure problem; (5) What examiners look for in RC-N past-due; (6) Instruction changes you missed this quarter; (7) RC-C loan code mistakes; (8) Reconciling the Call Report to your GL; (9) De-novo/first-time filer survival guide; (10) CECL & the Call Report. 3 diagnostic teardowns: anonymized edit-failure autopsy; a variance-narrative before/after; a schedule-by-schedule risk heatmap. 2 lead-magnet angles: "Call Report Burden Calculator"; "12-point Reporting Continuity Self-Assessment." 1 webinar: "Surviving Q-close when your Call Report person is gone." 1 outbound diagnosis template: a one-page "reporting continuity risk memo" personalized to a target bank's public UBPR profile.

41Lead magnet & waitlist plan

Lead magnet / free diagnostic: a "Reporting Continuity & Edit-Risk Scan" — the bank shares its last filed Call Report (public data via the CDR/UBPR) and receives a free automated report flagging peer-anomaly lines, likely edit-risk schedules, and a continuity-risk score. This captures the exact pain signal (which banks are shaky) and demonstrates the engine. Waitlist CTA: "Reserve your Q[n] preparation slot." Conversion path: scan → 20-min continuity consult → scoped single-quarter pilot → annual retainer. What the buyer receives before paying: a genuinely useful risk scan built from their own public filing — trust through demonstrated competence.

42Warm GTM plan

Work diagnostic/calculator users, webinar attendees, and association introductions first. Offer a free scan review call, then a single-quarter pilot at a modest onboarding price with the amendment-free standard. Prioritize banks showing trigger events (recent preparer departure via LinkedIn, prior amendment in public filing history, upcoming core conversion).

43Targeted outbound plan

Build a list of FFIEC 051 banks $100M–$1.5B from public FDIC/UBPR data. For each, auto-generate a one-page continuity risk memo from their public filings and lead outbound (email + LinkedIn) with the diagnosis, not a pitch. Sequence: memo → scan offer → consult → pilot. Personalize on asset size, recent variance anomalies, and any public amendment history.

44Answer-engine / search visibility plan

Preparers constantly search for specific edit-check codes, schedule instructions, and "how to report X on the Call Report." Publish authoritative, well-structured guides (edit-check glossary, schedule-by-schedule how-tos, instruction-change summaries) optimized for both classic search and answer engines (ChatGPT, Perplexity, Gemini) so Callright surfaces when buyers research the pain. Include clear entity/FAQ structure and cite primary FFIEC/NCUA sources. Capture visitors with the free scan.

45Pilot design & early-demand-trap mitigation

First cohort: 5 FFIEC 051 banks, single-quarter pilots. Cap: hard stop at 5 until intake, evidence requirements, and edit-check coverage are hardened. Incentive: discounted onboarding + amendment-free standard for charter design-partners. Trap mitigation: pilots are learning labs, not custom-work farms — anything outside the productized 051 scope is logged as future roadmap, not delivered ad hoc. A pilot proves the wedge only when it hits 0 edit failures, 0 amendments, and on-time delivery.

46Early-access feedback flywheel

Every filing generates structured feedback: which mappings the reviewer corrected, which edit failures recurred, which variance narratives the bank edited, and any examiner reaction. Product feedback (recurring patterns) becomes new mapping rules, edit heuristics, and narrative templates; one-off custom asks are triaged to roadmap. Weekly review of correction logs during pilots. A correction is "productized" when it becomes an automated rule/check rather than a manual reviewer step.

47Build-before-scale checkpoints

• After 5 banks: harden intake checklists, required-evidence lists, and edit-check coverage. Do not add banks until a filing runs input-complete→delivered without ad-hoc fixes.
• After 10 banks: harden SOPs, exception queues, reviewer checklists, and delivery templates; measure reviewer minutes/filing.
• After 20 banks: pause new pilots until COGS, rework rate, escalation rate, and cycle time are measured and within target. Acceptable temporary manual workarounds: novel-core mapping by hand. Unacceptable (signals non-scalable): recurring manual re-keying or repeated same-class edit failures.

487-day launch plan

1. Recruit a fractional reviewer with community-bank Call Report experience (ideally ex-examiner/CPA).
2. Stand up the intake portal, edit-check checklist (built to current FFIEC specs), and a working mapping template.
3. Build the free "Reporting Continuity & Edit-Risk Scan" from public CDR/UBPR data.
4. Publish the offer page + 2 cornerstone posts + the burden calculator.
5. Line up 3 warm intros via a bankers' association or CPA-firm contact; book 5 scan calls.

4930-day launch plan

1. Run 20+ free scans; convert 3–5 to single-quarter pilots for the next filing deadline.
2. Deliver first pilots semi-manually; capture every mapping and exception into SOPs.
3. Ship 8–10 content pieces; launch the continuity webinar.
4. Sign one referral partner (bankers' bank or bank CPA firm).
5. Instrument metrics: edit-failure rate, cycle time, reviewer minutes, conversion.

5090-day launch plan

1. Hit 5–8 paying banks on quarterly retainers; achieve 0 amendments across delivered filings.
2. Automate mapping + edit-checks to cut reviewer time toward ~65 min/filing.
3. Add NCUA 5300 and FFIEC 041 as second/third form types.
4. Formalize the amendment-free standard and vendor-risk-management packet for exam files.
5. Reach checkpoint gates before expanding past 10 banks.

51Metrics & KPIs

52Risks & mitigations (summary)

The dominant risks are (a) an errant filing causing a bank a finding/penalty (mitigated by dual machine+human gates and the amendment-free reserve), (b) bank procurement/vendor-risk friction (mitigated by a ready exam-file packet and starting with referral trust), and (c) deregulatory simplification reducing complexity (mitigated by the fact that the filing persists and change itself drives demand). Full register below.

53Exhaustive risk register

54What could kill this

The three most plausible killers: (1) a serious error on a client filing early on that destroys the accountability brand before referral trust is built — hence quality is over-engineered from day one; (2) inability to clear bank vendor-risk-management gates at scale — hence the exam-file packet and SOC 2 program are launch priorities; (3) a dominant incumbent bundling a genuinely good managed service — hence speed, small-bank focus, and partner-friendly positioning. A slower, non-fatal threat is deregulatory simplification, which erodes complexity but not the mandate.

55Go/no-go reasoning

The candidate clears every evidence-threshold requirement: a clearly identified buyer (small-bank CFO/controller), a painful and specific mandatory problem (quarterly Call Report with CMP exposure and no staffing bench), verified existing spend (software + CPA outsourcing + internal FTE), verified active demand and labor crisis, competitor/budget validation, a credible reason to win (productized AI outcome + accountability in an unoccupied slot), a narrow MVP wedge (FFIEC 051), a service-first delivery path (semi-manual first 3), no unresolved fatal regulatory/legal blocker (preparation not attest; bank certifies), a credible 55%+ margin path, and a believable distribution path (associations, bankers' banks, CPA white-label). No fatal disqualifier is triggered.

56Final recommendation

57Sources

1. [S1] OCC Bulletin 2025-42 — RFI on Call Report regulatory reporting burden. occ.gov
2. [S1b] Federal Register — Request for Information: Streamlining the Call Report (Dec 1, 2025). federalregister.gov
3. [S2] FFIEC 031/041 Instructions for Preparation of Consolidated Reports of Condition and Income. ffiec.gov (PDF)
4. [S3] FDIC — Current Quarter Call Report Forms, Instructions & Related Materials. fdic.gov
5. [S4] 12 CFR § 308.132 — Assessment of penalties (Call Report CMPs). law.cornell.edu
6. [S5] NCUA — Reinstates Civil Money Penalties for Late Call Report Filing (eff. Jan 1, 2024). ncua.gov
7. [S6] NCUA — Quarterly Credit Union Data Summary 2025 Q3 (4,331 FICUs). ncua.gov (PDF)
8. [S8] ABA Banking Journal — CSBS: Regulatory burden falls hardest on community banks (Nov 2025). bankingjournal.aba.com
9. [S9] HEDA Global — The Talent Shortage in Community Banking (2026). hedaglobal.com
10. [S10] FRB Services — Report Preparation Software Vendor Contact List (Mar 2025). frbservices.org (PDF)
11. [S10b] DBI Financial Systems — EasyCall Report. e-dbi.com
12. [S10c] FedReporter — Bank/CU regulatory reporting software (SmartCall). fedreporter.net
13. [S11] FDIC FIL — Revisions to the Consolidated Reports of Condition and Income (2025). fdic.gov
14. [S12] NCUA — 5300 Call Report FAQs. ncua.gov
15. [S13] Charles River CFO — Banking / credit union outsourced finance (incl. Call Report completion). crcfo.com
16. [S13b] CliftonLarsonAllen (CLA) — Banks industry services. claconnect.com
17. [S14] SEC — Proposed amendments to permit optional semiannual reporting (2026). sec.gov
18. [S15] Foley & Lardner — Federal Court Vacates FinCEN Residential Real Estate Reporting Rule (2026). foley.com
19. [S16] Grand View Research — RegTech Market Size & Trends. grandviewresearch.com
20. [S16b] Precedence Research — RegTech Market. precedenceresearch.com
21. [S17] Houseblend — SEC/regulatory filing software platforms, pricing & compliance (pricing context). houseblend.io

Confidence labels: Verified = directly supported by cited sources; Inferred = reasoned from verified facts; Unverified = plausible, not independently confirmed (chiefly the modeled pricing/margin/SAM figures, to be replaced with pilot actuals). FDIC bank count (~4,462) is drawn from FDIC 2025 reporting as surfaced in research; treat as Med-High confidence pending direct FDIC BankFind confirmation.