Mortgage AI Voice Receptionist in Oregon

Lead trigger is the lead-vendor feed, the company's web form, or an inbound call. The agent dials or texts within ninety seconds with a warm opening that references the loan purpose. The qualification runs through loan purpose (purchase, refinance, cash-out), property type and approximate value, employment and income range, credit estimate, current housing and timeline, and any specific lender or program preferences (VA, FHA, jumbo, conventional). The agent gives a preliminary affordability framing based on the income and credit and books the loan officer's pre-approval call. The LO gets the discovery transcript and walks into a substantive conversation rather than starting from scratch. CRM or LOS write-back to Encompass, Calyx Point, Floify, or BNTouch. A typical purchase lead call sounds like this. A borrower named James submits a LendingTree form at 9:18pm on a Sunday after a weekend of looking at houses with his real estate agent and getting serious about pre-approval. The webhook fires within thirty seconds and the agent dials his phone. James answers expecting a robocall and is surprised by the warm professional tone. The agent opens by acknowledging he submitted a purchase inquiry on LendingTree and runs the discovery: loan purpose (purchase, looking at homes in the four to five hundred thousand range), property type (single family home in his state), employment situation (W-2 employee at a software company for seven years, base salary one hundred sixty-five thousand with annual bonus), credit estimate (he checked his Credit Karma score last week and it showed seven sixty), down payment plan (twenty percent from savings, has the funds verified), current housing (renting, lease ends in four months), and timeline (looking to be under contract within sixty days). The agent gives a preliminary affordability framing: based on his income and credit profile, he is likely to qualify for a four to five hundred thousand purchase comfortably, the monthly payment on a four-fifty loan at current rates would run in the high two-thousand to low three-thousand range plus taxes and insurance, and the brokerage can typically issue pre-approval within twenty-four to forty-eight hours of receiving documents. The agent books James for a Monday morning pre-approval call at 9am with the brokerage's senior loan officer, sends a confirmation text with the LO's name and contact info, and a documents-needed list (most recent two pay stubs, two months of bank statements, two years of W-2s, driver's license). Total time from form submit to confirmed appointment: ninety seconds plus thirteen minutes of conversation. The loan-purpose-specific qualification flow is the trade-specific intelligence that separates this from a generic lead-handling template. Purchase loan discovery focuses on the property target (specific home under contract versus generally shopping), the down payment source (verified savings versus gift versus liquidating investments), the timing pressure (existing contract with closing date versus general timeline), and the relationship to a real estate agent (working with one versus not yet, which affects the LO's coordination needs). Refinance discovery focuses on the current loan terms (rate, balance, type), the refinance objective (lower rate, cash-out, change loan type), the property valuation expectation, and any timeline triggers (rate-lock urgency, expiring ARM, planned home improvements). Cash-out refinance discovery includes the use of funds because that matters for loan-to-value considerations. VA loans require veteran eligibility confirmation. FHA loans require first-time-buyer awareness and program-specific limits. Jumbo loans require higher-touch handling because of the borrower profile. The prompt applies the right qualification logic for each loan type with the brokerage's specific program offerings baked in during setup.

The mortgage lead funnel is broken in most brokerages because loan officers handle their own intake calls and are usually mid-loan with another borrower. New leads sit in a queue, get a callback an hour or two later, and discover the borrower already started an application with a competitor. The brokerages that solved this hired dedicated loan officer assistants at thirty-five to fifty thousand a year, plus the lead-quality dependency on the assistant's skill. The agent gives every brokerage that capability at a fraction of the cost and consistent quality. The specific labor economics that drive this leakage are worth understanding because they explain why so many brokerages have accepted the loss. A loan officer assistant or LOA earns thirty-five to fifty-five thousand a year fully loaded. They handle lead intake, document collection, file management, and borrower communication during the loan process. The cost is meaningful and the volume per LOA is limited; one LOA can typically support two to three active LOs depending on the loan volume each LO produces. Staffing a dedicated speed-to-lead intake function requires either taking an LOA away from active loan support (which slows down the loans in process) or hiring additional LOAs (which doubles the labor cost). Neither solution scales well, which is why most independent brokerages have accepted the speed gap and focus on relationship-based business rather than lead-vendor competition. The agent breaks this constraint because it handles unlimited lead intake at flat cost without touching the LOA's loan-processing work. The second structural issue is the after-hours and weekend lead-arrival pattern that does not match LO availability. Mortgage leads arrive heavily in the evenings and on weekends because borrowers research mortgages outside of work hours. Loan officers typically work eight to seven weekdays with limited weekend availability. So the Sunday evening lead, the Tuesday after-dinner lead, the Saturday afternoon lead, all of these go into the next-morning queue at most brokerages. By the time the LO calls back Monday morning, the borrower has already been contacted by Rocket Mortgage's 24/7 call center, Better Mortgage's evening shift, and two other independent brokerages who happened to be working late. The borrower has emotionally committed to whoever did the best first call. The agent solves this by handling the after-hours intake within ninety seconds at any hour, which means the brokerage captures the leads that previously went cold over the weekend. The Monday morning Lo wakes up to a fully qualified pre-approval call already on their calendar rather than starting from scratch with a queue of two-day-old voicemails.

Average mortgage origination commission for a broker runs one to two percent of the loan amount, or three thousand to ten thousand dollars per funded loan on typical residential mortgages. A brokerage receiving two hundred leads a month with a baseline ten percent funding rate is funding twenty loans for sixty to two hundred thousand in commission. Lifting funding rate to fifteen percent through better first-touch is ten extra loans, which is another thirty to one hundred thousand a month in commission revenue. The agent costs less than ten percent of one extra funded loan a month. Breaking the math down by loan type shows the variation that matters for understanding the revenue profile. Conventional loans on starter and mid-range homes (two hundred to four hundred fifty thousand loan amount) generate two thousand to five thousand in commission per funded loan. Conventional loans on mid-range to upper-mid homes (four fifty to seven fifty thousand) generate four thousand five hundred to ten thousand. Jumbo loans above the conforming limit (seven hundred fifty plus depending on metro) generate seven thousand five hundred to twenty thousand or more depending on size and pricing. FHA and VA loans typically generate commissions similar to conventional but with different pricing structures. Refinance loans generate similar commission patterns to purchase loans. So one captured lead converted to a funded loan varies from two thousand on small starter loans to twenty thousand or more on jumbo refinances, which makes every captured appointment economically significant. The lifetime customer value math is unique in mortgage because borrowers refinance multiple times across the life of a mortgage and the original lender of record has a meaningful advantage on the next refinance. The average homeowner refinances two to three times during the typical thirty-year mortgage tenure, and the lender that originated the previous loan captures the refinance about thirty to forty percent of the time when rates drop. So one well-captured purchase lead is not just one funded loan, it is roughly two to three funded loans across the borrower's ownership of that property. The referral chain is also strong because mortgage clients refer friends and family members who are shopping for loans; satisfied borrowers refer one to two friends in their first three years. Lifetime gross commission from one well-captured purchase lead routinely exceeds eight to twenty thousand dollars across refinances and referrals. The agent's recovery of leads compounds across this entire downstream pattern, which is why mortgage brokerage owners typically renew the retainer indefinitely once they see the year-over-year funded-loan numbers.

Vapi assistant tuned for mortgage discovery, with the loan-purpose-specific qualification branches, the affordability framing logic, and the LO routing rules. n8n workflow connecting to lead-vendor feeds and to the LOS or CRM. SMS confirmation and pre-call reminders. Knowledge base for common borrower questions (what credit score do I need, how much down payment, what documents will you need). Setup guide for lead-vendor connections, prompt customization to the brokerage's licensing footprint, and the affordability framing tuning. The LOS and CRM integrations ship for the major mortgage industry platforms. Encompass (Ellie Mae) has clean integration through their developer API with full loan-record creation capability. Calyx Point integrates through their API with slightly different setup steps. Floify has integration through their developer endpoints with strong document collection coordination. BNTouch is well-supported for the LO marketing CRM workflow. Surefire is supported for brokerages using it for marketing automation. Velocify integrates for brokerages running heavy lead-vendor programs. For smaller brokerages on simpler systems (a CRM plus the LOS) the template includes a basic integration that handles the workflow with full lead-to-LO handoff. The lead-vendor integration list covers LendingTree, Bankrate, Zillow Mortgage, Credit Karma, NerdWallet, and other major mortgage lead sources with webhook listeners that fire within thirty seconds of lead generation. The Vapi system prompt is the highest-value piece of the template and the part most resistant to commoditization. It includes the professional consultative tone that mortgage borrowers respond to (mortgage is a significant financial decision and borrowers want to feel they are being treated competently rather than rushed into an application), the bill-anchored affordability framing that creates immediate value, the qualification flow that captures everything the LO needs without making the borrower feel processed, the explicit guardrails against quoting specific rates or committing to specific pre-approval amounts before underwriting review (which is a regulatory requirement), the state-licensing logic that politely declines borrowers outside the brokerage's licensed footprint, and the compliance disclosure language that meets equal-housing and NMLS requirements. The prompt is the result of about four hundred test conversations across actual deployed brokerage accounts, refined against the conversational patterns that produce the highest lead-to-funded conversion. The rate-quoting guardrail is particularly important because quoting specific rates without licensure is a regulatory violation, and the prompt enforces this strictly.

Oregon has 4 million residents distributed across major metros including Portland, Eugene, Salem, Gresham, and Hillsboro. Oregon's CCB registration is the basic trust signal. Portland metro is shifting toward electrification under state climate policy. Recent heat events have reshaped AC adoption assumptions. The seasonality of mortgage work in Oregon is the single biggest factor that shapes how this ai voice receptionist actually performs in the market. Pacific Northwest patterns: extended rain season in western Oregon. Eastern Oregon continental. The template's qualification logic, dispatch rules, and conversation flow are tuned to handle these patterns rather than forcing the agency operator to customize from scratch. Shops that deploy this in Oregon markets see the seasonality framing show up in the conversations from the first call. Regulatory framework for mortgage brokers in Oregon varies at the local level rather than statewide, which is worth understanding because licensing references in customer conversations need to match local jurisdiction. The agent template handles this correctly by deferring licensing-specific questions to local context rather than asserting state-level rules that may not apply.

A day. The most important conversation with the broker is about the affordability framing logic: how aggressive to be with preliminary numbers, what license states the brokerage operates in, and how to handle borrowers outside the footprint. The LOS integration is the technical bottleneck. Encompass has the cleanest API. Floify is straightforward. Test by simulating both a purchase lead and a refinance lead. Agency operators in mortgage charge twelve hundred to twenty-five hundred for setup and five hundred to a thousand a month. The gotchas worth knowing before you go live are predictable but worth flagging. First, the state licensing footprint needs to be configured correctly because the brokerage cannot operate outside its NMLS-licensed states, and the agent must decline leads from non-licensed states cleanly without continuing the discovery. Second, the affordability framing math needs to be conservative because aggressive preliminary numbers that the formal underwriting cannot support create disappointment and erode trust; the agent should always frame numbers as preliminary ranges rather than commitments. Third, the lead-vendor integrations need to handle different data formats and missing fields gracefully because different vendors send leads with different completeness levels. Fourth, the disclosure language needs to match the brokerage's compliance requirements and any state-specific rules; the standard equal-housing language and NMLS identification are required but some states have additional requirements that must be configured. The ongoing tuning, if you want to do it, focuses on the qualification flow and the affordability framing calibration. Pull conversation transcripts weekly for the first month and look for patterns where the agent could have done better: a qualification question that the LO found insufficient, an affordability estimate that turned out to be too aggressive or too conservative, a borrower who ended the call before booking the pre-approval. Common findings include tightening the credit-tier conversation so borrowers in lower-credit brackets are routed to the right program specialist, refining the timeline questions to capture the actual urgency rather than aspirational language, adding scripts for the specific objections borrowers raise (concerns about closing costs, concerns about the appraisal, concerns about employment verification for self-employed borrowers), and adjusting the rate-environment language as the market changes. After about ninety days the prompt is well-tuned for the specific brokerage and ongoing tuning becomes optional.