Another dashboard does not automatically create a better workday. In Episode 4 of Settling Scores, cohosts Abdel Khawatmi of PRMG Got Mortgages and Eric Levin, cofounder of Model Match, bring AI adoption down to a manageable starting point: understand your business, review your calendar and turn useful information into scheduled action.
Episode originally published April 24, 2026. Archive recap added October 24, 2026.
Watch Settling Scores Episode 4: Data Adoption in AI on YouTube.
Key takeaways: information needs a next step
- Start with an honest audit of your current business and calendar.
- Choose one repeatable workflow before building a larger technology stack.
- Compare your business with relevant peers, not an unrelated production model.
- Turn recommendations into reviewed, realistic calendar blocks.
- Use reminders and weekly reviews to check execution, not just intentions.
The adoption problem is not a shortage of tools
Abdel describes experimenting with custom GPTs, websites and shared workspaces for his originators. His examples include Trello boards, transaction tasks and an MBS Highway connection through Zapier. The goal, as he explains it, is to reduce the need to move among disconnected tools and make information easier to use.
Eric raises the practical objection: most loan officers will not build that entire environment themselves. Even someone interested in AI must first invest time learning how to create the efficiency. A sophisticated demonstration can inspire an audience without giving them a starting point they can actually repeat.
That tension shapes the episode. Abdel’s extensive experimentation provides context, but the recommended first step becomes deliberately smaller. Before designing another platform, identify a routine that already exists and make it easier to understand and execute.
Borrow the clarity of a well-organized agenda
The conversation opens with Abdel’s account of a Washington, DC advocacy trip. He praises the organization provided by First Home IQ and the clear schedule for participants: where to go, whom to meet and what would happen next.
Eric connects that experience with an originator’s workday. A useful system supplies a cadence rather than another decision to make. The participant does not need to reconstruct the entire plan before taking the next step.
For a mortgage team, a suggested equivalent is knowing when to prepare for consultations, review outstanding commitments and follow up with referral partners. The useful design question is simple: does the workflow make the next appropriate action clearer? An agenda that people understand can be more helpful than a dashboard that leaves everyone to interpret it.
Audit the business you actually have
Abdel’s starting point is the calendar. Eric adds an explicit business audit: look at what you are doing now before asking technology to improve it. Otherwise, an assistant may optimize an incomplete picture.
A suggested audit can use one recent working week. Label scheduled consultations, preparation, partner conversations, administrative work and content creation. Then note meaningful work that happened but never appeared on the calendar. An empty block is not necessarily unused time, and a crowded calendar is not proof of productive activity.
For example, an originator may believe that partner follow-up happens every afternoon. The review might show that those blocks repeatedly became file-administration time. The useful finding is not that the originator needs more notifications. It is that an important responsibility lacks protected time or adequate support.
Keep the review descriptive before making it judgmental. What happened? What was completed? What was postponed? Those questions create a more reliable starting point than asking AI to assign a success score without explaining the work.
Use relevant comparisons, not an imaginary ideal
Eric discusses using Model Match to examine an originator’s business and compare it with the market. He emphasizes context, including geography, company type and referral base. A credit-union originator and a broker may work within substantially different structures.
During the April conversation, Eric says Model Match was building an AI tool to support comparisons inside the application. He presents that as development underway. The larger planning lesson is already useful: compare like with like and understand the business context behind a number.
Abdel suggests asking an assistant to evaluate a schedule against successful originator habits. Treat the resulting recommendations as hypotheses. An assistant should not be assumed to know another professional’s private schedule, complete production history or reasons for success. Verify external data and distinguish observed facts from generated suggestions.
A better comparison asks a narrower question: given the responsibilities and verified business information provided, where does the schedule fail to support the stated goal?
Connect only what the workflow needs
Abdel recommends connecting an AI assistant with a calendar and discusses Google Calendar, Outlook and a possible Zapier workaround for another calendar environment. The practical question for a team is which approved connection fits its actual account, permissions and working process.
Before connecting anything, use your organization’s approved tools and review the permissions requested. Calendar entries can contain borrower names, meeting links and confidential notes. A calendar audit does not require copying sensitive loan information into an unapproved assistant.
A suggested lower-risk starting exercise is a sanitized activity list: “consultation preparation, 30 minutes” or “partner follow-up, 20 minutes,” without identifying clients. Ask the assistant to organize those categories and identify scheduling conflicts. This tests the planning idea before expanding access.
The hosts also discuss paid AI access. When choosing a tool, evaluate the approved account’s actual features, memory settings, integrations and automation controls. The useful purchase decision begins with the workflow you need, rather than assuming every subscription will behave the same way.
Turn the recommendation into calendar time
The next step is execution. Abdel describes asking for a schedule in ICS format so it can be imported into a calendar. He gives the example of creating a baseline schedule for a business-development manager, providing something to return to when other assigned work is complete.
That is a useful distinction: a baseline is not an inflexible script. A loan officer still needs room for active files, changing appointments and conversations that require more attention than expected.
Here is a suggested prompt adapted from the episode’s approach: “Using these activity categories and available hours, draft a weekly schedule with consultation preparation, partner follow-up and administrative review. Preserve existing appointments, include buffers and explain your assumptions. Show me the plan before preparing an ICS file.”
Review the result before importing. Check dates, time zones, recurring events and conflicts. Test a small set of blocks first. A calendar import is only useful if it creates a workable schedule rather than duplicates and cleanup work.
Build a review rhythm that checks reality
Abdel’s next recommendation is a recurring rhythm: a morning preview, an end-of-day recap and a Friday comparison with the previous week. He describes using a project and scheduled tasks to support that habit.
Keep the planning workspace separate from the automation check. Naming a project does not establish that a reminder has been scheduled successfully. Where an approved tool supports recurring tasks, confirm the timing and test delivery before relying on it.
The daily review should distinguish scheduled work from completed work. A calendar can show that a consultation was planned; it may not establish that it occurred or that the promised follow-up was delivered. Add a brief completion note through an approved process.
A useful Friday question is: “Which commitments were completed, which moved, and what one scheduling change should we test next week?” That produces a specific adjustment rather than a broad motivational summary. A review earns its place when it helps the following week work better.
A suggested local application for Ocean and Monmouth County teams
For a team working in Toms River, Eatontown or surrounding communities, the framework can be applied using the team’s own responsibilities and observed scheduling problems. Begin with what is actually getting in the way, rather than an assumed local production benchmark.
For example, an originator with appointments in different locations could protect preparation time and realistic travel buffers. A teammate handling an educational session could reserve time for the promised attendee resources and individual follow-up. These suggested planning examples connect organization with personal service.
Measure the immediate workflow first: fewer missed preparation blocks, fewer scheduling conflicts or more promised follow-ups completed. If the process improves, ask what made the difference. Consider reporting periods, pipeline timing and other business changes before drawing conclusions about production.
A suggested five-step adoption checklist
The hosts close by condensing the discussion into five ideas. This suggested implementation checklist adapts that framework for a manageable first trial:
- Know your starting point. Audit a real week and identify one recurring obstacle.
- Choose an approved connection. Review access and begin with the minimum necessary information.
- Ask for a grounded assessment. Supply relevant context and require assumptions to be identified.
- Schedule one change. Review the proposed blocks before importing or repeating them.
- Review execution. Confirm reminders work and compare completed commitments each week.
The objective is not to replicate Abdel’s entire technology stack. It is to make one useful behavior easier to repeat, then decide whether the improvement justifies expanding the system. A smaller process that people understand can provide a better foundation for the next change.
Frequently asked questions
Who hosts this Settling Scores episode?
Cohosts Abdel Khawatmi of PRMG Got Mortgages and Eric Levin, cofounder of Model Match, discuss turning business data and AI tools into practical daily action.
What is the first AI adoption step recommended in the episode?
Start by understanding your current business and calendar. Review how time is actually used before asking an assistant to recommend a different schedule.
Do I need to build a custom platform to begin?
No. Eric specifically pushes for a smaller starting point. A focused calendar review and one repeatable action can be more manageable than recreating Abdel’s broader technology setup.
How was Model Match discussed in April 2026?
Eric discussed auditing an originator’s business and comparing it with relevant market peers. He also described an in-application AI comparison tool as being built during that conversation.
Can AI reliably tell me how top originators spend their day?
Not without reliable supporting information. Treat generated comparisons as suggestions, verify external claims and avoid assuming the assistant knows another professional’s complete schedule or business circumstances.
What is the purpose of an ICS file?
An ICS file can transfer proposed events into a compatible calendar. Review dates, time zones, recurrence and conflicts before importing, and test a small schedule first.
What should a weekly AI-assisted review measure?
Compare completed commitments, postponed work and scheduling obstacles. Calendar entries alone do not prove completion, and a weekly review should produce one practical adjustment.
Should borrower information be included in a calendar audit?
Use approved systems and minimum necessary information. A planning exercise can often begin with sanitized activity categories rather than borrower names, financial documents or confidential notes.
How can I contact Got Mortgages?
Call 973-536-0276 or visit GotMortgages.com to begin a conversation with PRMG Got Mortgages.
Make the next action easier
The strongest takeaway is modest: understand the work, choose a useful adjustment and check whether it happened. Better information matters when it supports preparation, thoughtful conversations and reliable follow-through.
To talk with PRMG Got Mortgages, call 973-536-0276, visit GotMortgages.com, contact the team or build your mortgage plan.
Explore more Mortgage Intel and Abdel Khawatmi’s profile. Explore more conversations at the Got Mortgages Podcast website.
Author: Abdel Khawatmi with PRMG Got Mortgages | 973-536-0276
This article is general mortgage information only. It is not a commitment to lend, a rate quote, or legal or tax advice. Loan approval, terms and pricing depend on a full application, credit review and property details, and market conditions change. Talk with a licensed loan officer and your own legal or tax professional about your situation. NMLS #1712023 | PRMG NMLS #75243.