Pavan contrasts a mortgage-industry cost figure he attributes to the Mortgage Bankers Association with Sun West's internal operating cost. He says Sun West was at roughly $300 per closed loan three years earlier and estimates its current cost at $100 to $150, excluding out-of-pocket items such as credit reports.

Pavan describes the automation savings as massive
He says Sun West had reached about $300 per closed loan three years earlier
He estimates the later operating cost at $100 to $150 per closed loan
His estimate excludes external costs such as credit reports

“three years ago we were at about 300 bucks a Clos loan and we've made it more efficient than that”
Pavan Agarwal · 21:00

“my guess we' be between 100 and 150 a closed loan”
Pavan Agarwal · 21:00

#mortgage costs#operating efficiency#automation#sun west

Explainer· 5

Explainer08:00

What Angel AI Does With a Mortgage File

A loan officer can scan borrower documents into Angel AI and focus on the customer relationship instead of paperwork. Pavan says the system reads documents, fills the loan application, identifies missing evidence, flags anomalies such as large deposits or varying overtime, checks calculations for fraud, and adds underwriting conditions.

The loan officer submits source documents rather than manually completing the application
Angel AI extracts data and completes the 1003
It identifies missing explanations and supporting documents
Cross-validation helps catch compliance problems and fabricated documents
It underwrites and adds conditions automatically

“drop it into Angel Ai and then I'm done”
Pavan Agarwal · 09:00

“it looks at every single data point on the on the documents and and because of that it's it also catches fraud”
Pavan Agarwal · 09:30

#angel ai#mortgage automation#underwriting#fraud detection

Explainer10:30

Angelistas Pair AI Instructions With Human Empathy

Sun West assigns loan officers a dedicated bilingual Angelista who follows up with borrowers, realtors, or title companies for information identified by the AI. Because Angel AI supplies what is needed and why, the human agent can concentrate on explaining the request, handling objections, and building trust.

The Angelista team operates from a bilingual call centre in Puerto Rico
A dedicated Angelista is assigned to each loan officer account
Angel AI provides detailed instructions for each follow-up
The agents receive weekly communication and objection-handling coaching
Humans focus on emotional intelligence and the customer relationship

“the AI tells them what to do they follow the instructions”
Pavan Agarwal · 11:30

“all they have to do is focus on understanding the customer building a relationship with the customer”
Pavan Agarwal · 13:00

#human in the loop#customer service#angelistas#empathy

Explainer13:30

How Angel AI Compresses Mortgage Processing Into One Hour

Pavan says Angel AI completes the 1003 from documents, performs verifications and lookups, and sends disclosures automatically. Work normally performed by a loan officer assistant or processor is therefore completed within the first hour, with OCR and cross-data validation reducing manual entry errors.

Documents become the source for the completed 1003
Verifications and external lookups run automatically
Customer disclosures are sent automatically
OCR replaces manual document reading and data entry
The workflow targets completion of processor tasks within the first hour

“the 103 gets filled from the documents”
Pavan Agarwal · 13:30

“everything that that what normally an Loa or processor does is done within the first hour”
Pavan Agarwal · 15:00

#loan processing#ocr#automation#disclosures

Explainer40:30

Mortgage Professionals Are Splitting Into AI Rejecters and Adopters

Pavan describes mortgage professionals moving from denial and criticism toward a split response to Angel AI. Some avoid it because they fear for their jobs and future, while others say it transforms their business; he notes that both reactions occur among large producers, whereas realtors tend to adopt it readily because they view it as helpful rather than competitive.

Initial reactions included disbelief and dismissal
Criticism followed as users began testing the system
Some loan officers now reject it because they feel threatened
Others ask for more because it expands what they can do
Realtors tend to see it as a business-generating aid

“this is going to take away my job it's going to take away my future”
Pavan Agarwal · 41:30

“another group which is give me more”
Pavan Agarwal · 42:00

#ai adoption#mortgage professionals#job fear#realtors

Explainer43:00

Why Angel AI Was Designed Around One Button

Pavan says the interface is deliberately sparse so using Angel AI feels like talking with a trusted friend. Inspired by Apple's uncluttered iPhone and Google's search page, he pushed engineers toward zero buttons and accepted essentially one send button, making natural conversation replace menus.

The intended experience is a comfortable conversation
Empathy training is an ongoing part of the product
The interface is purposely simple enough to surprise users
Apple inspired the uncluttered design
Google demonstrated that a single input can replace a crowded page

“I don't want any buttons If you could it with zero buttons do it”
Pavan Agarwal · 44:00

“Simplicity always wins”
Pavan Agarwal · 46:00

#user interface#simplicity#conversational ai#product design

Story· 3

Story05:30

The Four-Decade Foundation Behind Angel AI

Pavan traces Sun West's automation mindset to his father, who viewed mortgages as paperwork, calculations, and recurring patterns. After writing an AI program in 1985, Pavan concluded that data, storage, and computing were not yet sufficient, so he established an architecture for collecting data and building toward future automation.

His father repeatedly urged him to automate the mortgage business
Pavan wrote his first AI program in 1985
Early storage and computing limits prevented the larger vision
Sun West built a data and engineering foundation while waiting for technology to catch up

“the first AI program I wrote was back in 1985”
Pavan Agarwal · 06:00

“at least put an architecture in place so that we can start collecting data”
Pavan Agarwal · 07:00

#ai history#automation#data architecture#mortgages

Story18:00

The Angel AI Warranty Has Cost About Four Loans

After years of training and human correction, Sun West launched Angel AI in 2018 and introduced a warranty in 2019. Pavan says that if the system wrongly approves a loan, Sun West will still fund it and hold it on its balance sheet; he reports doing so about four times after more than 200,000 Angel AI transactions since 2018.

Early errors required humans to correct and tune the system
Angel AI officially launched in 2018
The warranty began in 2019
Sun West accepts the loan if Angel AI approved it incorrectly
Pavan reports about four balance-sheeted loans under the warranty

“if it got something wrong I'll eat that loan”
Pavan Agarwal · 19:30

“since 2019 to today I've had to balance she about four loans”
Pavan Agarwal · 20:00

#ai warranty#model accuracy#accountability#mortgages

Story28:00

How Childhood Poverty Became Entrepreneurial Optimism

Pavan remembers immigrating from poverty in India and experiencing even a one-bedroom American apartment as a luxury. He says he interpreted the period through childlike wonder rather than as a set of challenges, and believes that positive perspective contributed to his entrepreneurial resilience.

His family immigrated from poverty in India
American conditions felt luxurious relative to what he had known
He remembers excitement more than hardship
Surviving that starting point makes future adversity feel manageable

“I didn't see the challenge everything was always positive”
Pavan Agarwal · 29:00

“still I like to keep that same sense of wonderment and positive positive mindset”
Pavan Agarwal · 29:00

#immigration#resilience#optimism#entrepreneurship

Tool· 1

Tool15:30

Why Competing Loan Officers Use Angel AI for Free

Pavan says Angel AI is free to loan officers even when they work for mortgage banks that will never submit a loan to Sun West. Those users can upload applications and documents to obtain qualifying income, asset analysis, problem flags, and a set-up loan file without adding another assistant.

Access is available to loan officers outside Sun West
Users can rely on Angel AI to review income and assets
The system reports qualifying income and potential problems
Some loan officers describe it as a free assistant that saves hours

“I make the angel AI completely free and accessible to every to company to every company”
Pavan Agarwal · 16:00

“you you Sav me hours every day”
Pavan Agarwal · 17:00

#angel ai#loan officers#product access#productivity

Takeaway· 5

Takeaway02:00

Pavan Agarwal's 10-Minute Morning Workout

Pavan starts around 7, checks social media, posts his thoughts, and tells his team his plan for the day. He then uses small dumbbells for a short routine designed to raise his heart rate for 10 to 15 minutes rather than training for hours.

He communicates his daily plan to his team in the morning
His office workout uses small dumbbells
The routine aims to raise his heart rate for 10 to 15 minutes
He chose consistency over intensive bodybuilding-style training

“the idea is just to get the heart rate up for about 10 15 minutes”
Pavan Agarwal · 02:30

“if you just do that and eat right you're going to be um in good shape forever”
Pavan Agarwal · 03:00

#morning routine#fitness#consistency

Takeaway22:30

Why Pavan Treats Relationship-Building as the Core Success Skill

Pavan argues that competitive sport teaches children how to win, lose, take a hit, respond within boundaries, and interact with people in the real world. He connects those lessons to professional success, saying that understanding cultures and forming genuine connections creates a deeper understanding of any eventual customer.

Competitive sport exposes children to wins, losses, and interpersonal boundaries
Real-world interaction develops character differently from virtual interaction
Multicultural experience can reveal that people think differently
Customer understanding is necessary to deliver a meaningful product

“all success boils down to your ability to to build relationships and create connections”
Pavan Agarwal · 23:30

“if you don't understand the customer like really understand them you're never going to be able to deliver a product that they that means something…”
Pavan Agarwal · 24:30

#relationships#sports#customer understanding#culture

Takeaway25:00

The Parenting Lesson Pavan Places Above Everything Else

Pavan says children learn most by watching their parents, particularly how their mother and father treat each other. He also involves his children in his work and travel so they can observe effort, understand why he is away, and see that money does not simply appear.

Parental behaviour teaches more than verbal instruction
Mutual respect between parents is the central example
Pavan lets his children observe his work and travel
Participation helps connect money with effort

“the kids learn the most by watching their parents”
Pavan Agarwal · 25:00

“let them observe right let them be a part of your life”
Pavan Agarwal · 26:00

#parenting#role modelling#respect#work ethic

Takeaway36:30

Pavan Calls Marriage His Most Significant Milestone

Asked for the greatest milestone in his career, Pavan answers that it was getting married and credits his wife with keeping him grounded. He connects that stability to listening to an inner calm before choosing what to do next, rather than making decisions from anger or excitement.

He married at about 26
He says he would be nowhere without his wife
Marriage helped keep him grounded and thinking clearly
He warns against decisions driven by anger or excitement

“getting married without I mean I got married young I was think was 26 um and without her I I nowhere”
Pavan Agarwal · 36:30

“that's when the worst decisions are made”
Pavan Agarwal · 38:00

#marriage#decision making#emotional regulation#success

Takeaway46:30

Why 15 Minutes of Meditation Became Pavan's Central Goal

Pavan says his personal, family, and business goals all return to making time for 15 minutes of meditation each morning. He recalls that life felt better when he maintained the practice and believes renewed discipline would help him make better decisions for customers while letting his children follow their own paths.

He wants to restore a 15-minute morning meditation habit
He remembers broad benefits when the practice was consistent
He does not want to prescribe goals for his children's independent lives
For the business, meditation supports better customer decisions

“what was missing in there was you know 15 minutes of meditation”
Pavan Agarwal · 47:00

“I got to get back on that diet of that 15 minutes of meditation every morning”
Pavan Agarwal · 47:30

#meditation#discipline#family#decision making

Frameworks in this episode

Everyone-Around-You Customer Model

Explainable High-Stakes AI Validation

Tombstone-to-Decision Alignment

Insights & moments

◆Hot Take· 1

Pavan Claims Sun West Closes a Loan for Under $200

✶Explainer· 5