Safe Rate's AI Mortgage Agent Scores the Loan Estimate in Under 30 Seconds

The Chicago startup, backed by $715,000, uses Google's Gemini to give instant quotes and has saved users up to $12,000 a year in interest.

About Safe Rate

Published

The average mortgage shopper might see a dozen different loan estimates before signing. Each one is a dense, three-page document filled with fees, rates, and fine print. Safe Rate, a Chicago-based fintech founded in 2018, runs those documents through an AI agent and returns a score from zero to one hundred in under half a minute. The company claims its benchmark is built on two million historical loans [Hacker News, June 2024].

It is a small wedge into a massive, opaque market. The bet is that speed and transparency, delivered through a chat interface, can pull consumers away from traditional brokers and direct lenders. Co-founders Shima Rayej and Dylan Hall, who met during their MBA program, built the platform to function as both a digital mortgage marketplace and an AI sales assistant [Safe Rate]. The tool, which leverages Google's Gemini and Document AI, promises personalized rates in 20 seconds without requiring a social security number upfront [Google Cloud, July 2024] [Safe Rate, 2025].

The Chat-Native Wedge

Safe Rate's product surfaces in three primary ways for consumers. A user can answer a few questions for an instant quote from its network, upload an existing loan estimate for the AI to try and beat, or submit a current mortgage statement to analyze refinance potential [Google, 2024]. The system is accessible via web chat, email, phone, and Slack, with plans to expand to RCS, SMS, and WhatsApp [Hacker News, June 2024].

For loan officers, the platform automates the pre-application sales process. The AI assistant can join live calls, listen in, and help price loans in real time while the human officer focuses on the client [Hacker News, June 2024]. The company says this lets officers "gain more leads, communicate more effectively, and win more loans" [Safe Rate]. It is a two-sided play: attract consumers with a self-service, data-safe experience, then feed qualified leads to a network of lenders.

The Team and the Technical Backbone

The founding duo brings a mix of product and technical depth. Shima Rayej, the Chief Product Officer, studied at MIT and is based in Chicago [LinkedIn, 2026] [RocketReach, 2026]. Dylan Hall, the CEO and Head of Mortgage Lending and Compliance, is a lecturer in the Masters Program in Computer Science at the University of Chicago and holds an active NMLS license [University of Chicago, 2026] [LinkedIn, 2026]. His background includes roles as a CTO, data scientist, and full-stack developer [LinkedIn, 2026].

Their technical stack is explicitly built on Google Cloud. The core AI mortgage agent is powered by Google's Gemini and Document AI models, a decision that provides robust infrastructure but also creates a clear vendor dependency [Google Cloud, July 2024]. The company claims its hyperlocal data covers rates, lenders, and ownership costs for over 80,000 places and 2,000 lenders [Safe Rate].

Traction and the Funding Picture

Public traction metrics are limited but pointed. The company states mortgage shoppers have saved between $100 and $12,000 per year in interest by using its platform [Safe Rate]. It also promotes a $2,000 credit toward a user's mortgage loan as an incentive [Safe Rate, 2026]. For now, its live lending operation appears focused; one listed partner, STATE FINANCIAL NETWORK, serves four states and offers conventional loans [Safe Rate, 2026].

Funding has been modest. According to third-party data, Safe Rate has raised approximately $715,000 across one round [Prospeo]. A separate source cites a $200,000 raise [CB Insights, 2026]. The lead investors for this capital are not publicly named, though the company has participated in the GET Seed Founder Program and is connected to the University of Chicago's Polsky Center for Entrepreneurship.

Role Name Background
Co-founder, CEO Dylan Hall Lecturer, University of Chicago CS; former CTO/data scientist; NMLS #1658740 [LinkedIn, 2026] [University of Chicago, 2026]
Co-founder, CPO Shima Rayej MIT alum; based in Chicago [LinkedIn, 2026] [RocketReach, 2026]

Where the Model Faces Pressure

The risks for Safe Rate are familiar for early-stage fintechs navigating a complex, regulated industry. The model depends on a few critical motions that remain unproven at scale.

  • Consumer trust. The promise of no sensitive data upfront is a powerful acquisition tool, but converting a casual rate shopper into a closed loan requires eventually collecting that data and guiding them through a full, compliance-heavy process. The final mile still involves a human loan officer.
  • Lender network growth. The value to consumers hinges on the breadth and competitiveness of the lender marketplace. A limited network could result in uncompetitive quotes, undermining the core "Beat This Rate" proposition. The company has not publicly named major lender partners.
  • Capital intensity. Mortgage lending is a balance-sheet business. With under $1 million in disclosed funding, Safe Rate's ability to scale its own lending operations or secure significant partnership commitments may be constrained compared to well-funded competitors.

The company's answer likely lies in its asset-light approach. By positioning primarily as an AI-powered marketplace and assistant, it aims to avoid the capital burdens of a full-stack lender. Its tools for loan officers are designed to generate revenue by improving their efficiency, not by carrying loan risk.

The Next Twelve Months

For a company founded in 2018, the next phase is about proving it can transition from a promising tool to a scalable business. Key milestones to watch include a named institutional funding round, the announcement of a major lender or enterprise partner for its B2B tools, and a clear expansion of its live lending footprint beyond its initial geographic focus.

The $715,000 in backing from the GET Seed Founder Program and Polsky Center is a start, but the mortgage market rewards scale. The question for Rayej and Hall is whether their AI-powered wedge,scoring a loan estimate in 30 seconds,is sharp enough to carve out a sustainable piece of the process before larger players decide to build or buy similar capability.

Sources

  1. [Safe Rate] Company website | https://saferate.com/
  2. [Google Cloud, July 2024] Safe Rate helps homebuyers and owners save thousands with AI mortgage shopping | https://www.youtube.com/watch?v=oviFAoFVZUg
  3. [Google, 2024] Safe Rate | Gemini API Developer Competition | https://ai.google.dev/competition/projects/safe-rate
  4. [Hacker News, June 2024] Show HN: SafeRate - AI chat-native mortgage lender | https://news.ycombinator.com/item?id=44749241
  5. [Crunchbase] Safe Rate - Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/safe-rate
  6. [LinkedIn, 2026] Dylan Hall - CEO and Co-Founder - Safe Rate | https://www.linkedin.com/in/dylan-m-hall/
  7. [LinkedIn, 2026] Shima Rayej - Safe Rate | LinkedIn | https://www.linkedin.com/in/shima-rayej-b1b98278/
  8. [Prospeo] Safe Rate funding data | https://www.prospeo.io/
  9. [CB Insights, 2026] Safe Rate funding data | https://www.cbinsights.com/
  10. [University of Chicago, 2026] Dylan Hall lecturer profile | https://masters.cs.uchicago.edu/
  11. [RocketReach, 2026] Shima Rayej contact information | https://rocketreach.co/

Read on Startuply.vc