AI Mortgage Workflows: Abdel Khawatmi’s Conversation-First System

AI can help mortgage originators improve conversations, not simply automate messages. In Part 1 of his appearance on HousingWire’s Power House podcast, Abdel Khawatmi explains to host Zeb Lowe how recorded calls, structured follow-up and application-triggered education support that goal. His central argument is that conversations are the most overlooked part of conversion tracking: counting leads and closings does not explain whether a client felt understood or knew what to do next. For Got Mortgages Insights readers, the practical lesson is to build technology around better communication rather than treat technology itself as the service.

Conversation originally published August 25, 2026. Recap added October 24, 2026.

Watch the original conversation on YouTube.

Key Takeaways: Measure the Conversation, Not Just the Outcome

Abdel’s six Cs are content, client, conversations, credit, contract and closings. He identifies conversations as the stage originators and agents tend to underestimate. A business can track inquiries, credit activity and completed loans while still missing why a promising relationship stopped moving forward.

His workflow gives each conversation several potential uses: a record for the team, material for coaching and a basis for educational follow-up. Those uses are related but distinct. A coaching assessment helps the originator improve; a client recap helps the borrower understand the discussion.

The broader operational lesson is to assign a trigger, an owner and a purpose to each tool. Abdel describes application notifications, team communication, loan comparisons and partner recaps as connected steps. The practical opportunity is not to buy every platform he mentions, but to remove the handoffs that leave people uninformed.

A Poor Homebuying Experience Became a Communication Standard

Abdel connects his entry into mortgages to dissatisfaction with his own property purchase. After graduating from Rutgers and working at a personal injury law firm, he saw an opportunity to provide a different experience through consistent communication and education. He briefly worked in both settings before committing to the mortgage business.

That origin story matters because it explains the purpose of his systems. The problem he wanted to solve was not a shortage of software. It was the experience of making a major financial decision without the clarity and service he expected. Technology became a way to make the preferred experience repeatable.

Zeb brings a complementary perspective from his own transition out of teaching. He initially helped with presentations for real estate professionals and first-time buyers. Both accounts place education close to the center of origination, rather than treating it as an optional marketing activity.

For a mortgage team, the practical step is to define communication standards before selecting tools. Decide what a borrower should understand after an initial discussion, who handles unanswered questions and how the next step will be confirmed. Those decisions give an automation a useful job.

Use the Six Cs to Locate Gaps Between Activity and Progress

The six Cs connect visibility to completed transactions without reducing the business to a lead count. Abdel explicitly includes hyperlocal activity within content, not just social media. That makes the first stage broader than posting: local educational outreach can also introduce a prospective client to the business.

At the other end, credit activity, contracts and closings are more readily counted. Abdel describes watching how credit activity can precede later closings. His observation is a pipeline-management insight, not a reason to pull credit unnecessarily or assume a fixed relationship between credit reports and funded loans.

Conversations sit in the less visible middle. An originator may know how many people were contacted without knowing whether explanations were too long, too brief, insufficiently empathetic or poorly matched to the consumer’s needs. Those are different problems, and more outreach alone does not necessarily address them.

A practical tracking approach is to pair stage counts with a small set of conversation observations. Record whether the main concern was identified, whether an explanation resolved it and whether a next step was agreed upon. Reviewing those observations alongside progression can reveal questions that raw production totals cannot answer.

Turn Recorded Calls Into Focused Coaching

Abdel describes recording consumer conversations through his calling and meeting systems. RingCentral supports his phone workflow, while Fireflies captures meeting discussions. Zapier helps route information to the team. He says the practice developed from recording partner meetings and training sessions into a broader communication process.

The coaching component depends on instructions. Zeb asks whether Abdel tells the AI what to examine, including excessive length or missed opportunities for empathy. Abdel confirms that the criteria are supplied through project instructions. The system is therefore being directed toward a defined communication standard rather than asked for an unexplained overall score.

A useful editorial adaptation would be to ask for one place where the originator answered clearly, one unresolved concern and one suggested improvement. Require references to the relevant exchange, and have a person assess whether the recommendation fits. A shorter call is not automatically a better call, particularly when a borrower needs more explanation.

Recording and processing mortgage conversations also requires deliberate controls. Abdel mentions an automated recording notice; teams should separately establish applicable consent requirements, approved tools, access permissions, retention practices and limits on sharing borrower information. A recording announcement alone should not be treated as a complete compliance program.

Make Follow-Up Easier to Understand, Not Just Faster to Send

One of the episode’s most distinctive workflows turns a meeting transcript into several educational formats. Abdel says his team uses NotebookLM to create a video overview, a slide deck and an infographic. He also describes a Claude project that condenses a partner conversation into a short, conversational message.

These outputs solve different communication problems. A concise message can identify the next action, while a visual overview can help someone revisit a complicated explanation. The useful principle is to choose a format for the recipient’s need rather than assume every meeting deserves every possible attachment.

Abdel discusses editing outputs to correct branding and name spelling, and to remove information he does not want included. That review is central to the workflow. Generated material can sound polished while misstating a detail, including an unnecessary disclosure or giving too much emphasis to a passing comment.

His business development manager has a defined role in preparing partner follow-up. That makes ownership as important as generation. An effective recap should distinguish what was discussed from what was decided and what still needs confirmation. For consumer communications, the same discipline helps prevent an educational summary from being mistaken for a final lending determination.

Use the Application as a Trigger for Shared Understanding

Abdel describes the application as the start of a structured education and communication sequence. His team receives an automated notification and prepares a Mortgage Coach total cost analysis, or TCA. He also describes using the MBS Highway pre-approval buyer journey to support ongoing education.

The contrast with his earlier process is revealing. Previously, the team downloaded individual educational resources, including real estate report cards and buy-versus-wait materials, and distributed them through its own CRM. A more integrated buyer journey changed how that work was assembled and delivered.

A team member also establishes a group text involving Abdel, the client and the real estate agent. If the referring agent is unclear, the team checks rather than leaving the relationship unresolved. This is a concrete example of a handoff rule: an application should not simply enter a queue without anyone establishing who needs to communicate.

For teams adapting this approach, shared communication should make coordination easier without exposing sensitive financial details unnecessarily. Educational comparisons should also identify their assumptions and be updated as relevant inputs change. The objective is a common understanding of the process, not a single channel containing every piece of borrower information.

Build Content Around Client Questions and Partner Participation

Abdel’s content operation extends beyond his own appearances. He describes an office podcast studio, an in-house videographer and originators bringing referral partners into educational conversations. The arrangement gives team members and partners a role in explaining topics, rather than requiring one person to supply every idea.

He also describes using audience retention to shape episode length. His team generally aims for focused conversations of roughly 25 to 30 minutes after observing where viewers dropped off. That is a decision based on his audience, not a universal rule for mortgage podcasts. The transferable habit is to examine attention and adjust the format.

The publishing rhythm includes a Monday market update and a Friday market-signals communication. For the latter, Abdel describes exporting information from Model Match and using a Claude project to reshape it for New Jersey consumers, addressing market developments, interest rates, inventory, supply and demand.

The editorial opportunity is to translate information into understandable questions without adding unsupported conclusions. A local update should make clear what period it covers and why the development matters to a buyer. Recurring formats are most valuable when they consistently answer real questions, not merely fill a calendar.

Evaluate AI Tools by Their Role in the Business

Abdel organizes his thinking around sales, service and support. He describes ChatGPT as more useful in his operational-support work, including custom workflows and connected scheduling tasks, while Claude has recently taken a larger role in marketing, branding and document-related work. These are his working preferences, not a universal ranking of models.

Zeb challenges the habit of layering new technology over legacy processes without rethinking the underlying work. That distinction runs through the interview. A transcript generator alone does not create better follow-up; someone must decide where the transcript goes, what gets produced and who checks it.

Abdel says he reviews technology spending with his revenue manager and questions whether tools can be removed. He believes the tools he retains are being used and supporting the business, but the discussion does not provide a quantified return-on-investment analysis. A practical evaluation should therefore look at actual use, duplicated functions, review burden and observable service improvements.

His most telling observation is that he is spending more time on the phone, not less. Automation supports his conversations rather than replacing them. Teams can test that principle directly: if a new workflow creates more administrative review without improving client contact or understanding, it may need redesign.

Extend Partner Service Without Overpromising AI Visibility

The conversation also reaches beyond loan files. Abdel describes experimenting with coding websites and apps, and using answer engine optimization, or AEO, site checks to identify potential issues for referral partners. He shares findings as a way to offer useful attention to their businesses.

Zeb frames early technology adoption as an opportunity to differentiate. The strongest practical version of that idea is modest: notice a specific problem, explain it clearly and offer an appropriate next step. A partner may value a useful observation even when fixing the issue is outside the originator’s core work.

A website assessment does not establish that an AI system will recommend a business. Crawlability, content quality and recommendation behavior are different questions. The relationship-building lesson is to be helpful within the limits of what the assessment actually shows, while keeping mortgage service at the center of the partnership.

Final Checklist: Pilot One Complete Communication Workflow

Define the gap first. Choose one recurring conversation, such as an initial buyer consultation or referral-partner meeting, and identify the problem: unclear next steps, inconsistent follow-up or missed client concerns.

Set the operating rules. Confirm approved recording and AI practices, assign a reviewer and specify what information must stay out of shared outputs. Decide how the team will handle missing or uncertain details.

Build one complete sequence. Connect the conversation record to a focused coaching review or a concise recap, then assign responsibility for delivery and any unanswered questions. Add visual formats only when they serve a clear purpose.

Review the result before expanding. Look at whether the recipient understood the next action, whether staff had to correct the output and whether the workflow made personal follow-up easier. Keep what improves communication; revise what merely produces more material.

Frequently asked questions

What are Abdel Khawatmi’s six Cs of conversion tracking?

They are content, client, conversations, credit, contract and closings. The framework connects awareness-building activity to completed loans while drawing attention to the conversations that influence movement between stages.

Why does Abdel identify conversations as the most overlooked stage?

Teams can count clients, credit activity and closings more easily than they can evaluate communication quality. Abdel focuses on whether a discussion was too long or too short, showed enough empathy and gave the consumer the right level of explanation.

How does Abdel use AI to review mortgage conversations?

He describes capturing phone and meeting conversations, then using project instructions to guide AI feedback. The instructions identify communication qualities to evaluate, such as empathy and length. Human review remains important because an AI assessment may misunderstand the context.

What does his team create from meeting transcripts?

Abdel describes using NotebookLM for video overviews, slide decks and infographics, plus Claude for brief conversational summaries. He also discusses correcting names and branding and removing information that should not appear in the finished material.

What happens after a mortgage application enters his workflow?

His team receives an automated notification, prepares a Mortgage Coach total cost analysis and uses a pre-approval buyer education journey. A team member also establishes communication among Abdel, the client and the real estate agent, checking the referral relationship when necessary.

How does Abdel divide work between ChatGPT and Claude?

In this discussion, he describes using ChatGPT more for operational support and Claude more for marketing, branding and document work. That division reflects his workflow preferences; teams should evaluate tools against their own approved uses and review requirements.

How does his team create local market education?

Abdel describes Monday market updates and Friday market-signals communications. For the latter, the team exports Model Match information and uses a Claude project to reshape it for New Jersey consumers, covering market conditions, rates, inventory, supply and demand.

What is a practical first workflow for a smaller mortgage team?

Start with one recurring consultation and one useful output, such as a reviewed recap identifying the client’s question and agreed next step. Establish approved data handling and assign an owner before adding multiple formats or automated delivery. Evaluate whether the recap actually reduces confusion.

To talk with PRMG Got Mortgages, call 973-536-0276, visit GotMortgages.com, contact the team or build your mortgage plan.

Explore more Got Mortgages Insights 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.