Perspectives on AI, technology, and compliance transformation to help you move faster, smarter, and with more clarity.
featured insights
Tela Mathias, CTO & Chief Nerd at PhoenixTeam has been named a 2026 HousingWire Vanguard, marking her third consecutive year receiving the recognition. The annual award honors executives whose leadership, vision, and influence are helping shape their organization and the broader housing economy.
Tela’s recognition reflects the growing reach of her work across AI innovation, industry education, and workforce readiness. Tela’s work continues to shape how the mortgage industry approaches artificial intelligence, from moving practical AI solutions into real-world use to helping organizations think more intentionally about adoption, governance, and the role people play alongside the technology. Her leadership has helped advance PhoenixTeam’s work across AI innovation and education while creating new opportunities for mortgage professionals to build the skills needed to apply AI responsibly.
Her influence reaches beyond individual AI initiatives. At a time when AI is changing the industry so quickly, Tela’s work has continued to evolve with it. Her work is helping push the industry toward a more connected way of thinking about innovation, one that considers not only what technology can do, but where it can create lasting value for the business and the people behind it.
Read more about Tela’s recognition and the work behind it: https://www.housingwire.com/winner-profile/2026-vanguard-tela-mathias/
See the full list of 2026 HousingWire Vanguards: https://www.housingwire.com/articles/announcing-the-2026-housingwire-vanguards/
featured insights
As ChatGPT-era automation scales, leaders are pushed to redesign value beyond faster workflows.
I was recently in the middle of nowhere Utah with my family, and it was the first time in a very long time that I set my intentions on being present and being unplugged. My entire workday, which spans anywhere from four to 16 hours, is spent in the guts of artificial intelligence and how to use it to change the mortgage industry for the better. This runs in stark contrast to the experience of being present and connecting with my family. There is certainly very little that is artificial about the national parks in Utah.
I came back refreshed and reflective. Hopeful and maybe a little bit scared. Artificial intelligence, inasmuch as it is how I make my living, has enabled these precious human moments. There is a clarity that comes when we remove the noise of the day-to-day and create real human meaning. When we watch the stars appear, when we center on each other. There is something really important that happens when we discover something subtle and special about the humans in our lives.
Having just turned 50, it’s frankly hard to believe I’ve worked for 28 years, perhaps harder yet to believe that all of those years have been in mortgage technology. In November of 2022, artificial intelligence changed everything with the mass (free) availability of ChatGPT.
We didn’t really know it then; my own hard personal and company pivot wouldn’t come for another year, but the nature of work was altered forever. It would take a few years for generative artificial intelligence (genAI) to move beyond the “fun toy” stage, but anything that can be read, retyped or reconciled can now be automated.
And yet the average cost to originate a loan has increased. The average cost to service a loan has remained flat. This tells me that the transformational advantage that will come is limited to a few, if any.
It’s still bogged down in gaining literacy that leads to fluency, the challenges of scale, navigating safety and unknown regulatory constraints. And people, it’s really bogged down in the humans that have to understand it, figure out how to deploy it, learn to control it and find those places where real value can be realized.
Real mortgage transformation will not come from making the current manual processes faster. Anyone can do that. Frankly, with the recent advancements in agentic engineering, genAI correctly implemented can already do that autonomously. You won’t find the future of mortgage on your use case list, even though that is an important asset that will be heavily scrutinized by regulators.
Point-based solutions are a great place to start to build literacy and save some dollars. The “winners” of the AI race in mortgage will be the ones that reimagine mortgage now that the previously impossible is possible.
When anything that can be automated is automated, what remains? When getting a mortgage moves from a largely linear, digitized paper process to a one-shot straight-through decision, what’s left? When the homebuyer of the future has a personal financial agent instantly finding, negotiating, and closing their loan – what’s the value of a mortgage company?
It is those subtle special human moments that will matter, not the automated process. It will be the extent to which we have anticipated needs, we have found unspoken value, we have met that hopeful or fearful human exactly where they are at exactly the right time (that big promotion, the job loss, a parent’s death). Creating a mortgage organization that can do this requires discerning judgment, aspiration that can be implemented and real human creativity – none of which can be automated.
This reimagined ecosystem will leave a whole new kind of work for humans. Who are these humans? What do they do? What are the jobs? What’s that better future that will emerge, and how do we make it safe?
Over the next six months, in which time the pace of change will continue to consume us, I hope to answer some of these questions. I’ll define the future of mortgage. I’ll describe what’s left when anything that can be read, retyped or reconciled is automated.
We’ll close the year defining the minutes that matter and figuring out how to make more of them. Then we will center on the people and the jobs of the future. We’ll close out the series with the specific things mortgage leaders can do to realize this new future. I hope you will engage in and steer this conversation with us through comments or by reaching out directly.
Tela Gallagher Mathias is PhoenixTeam’s Chief Technology Officer and CEO of Phoenix Burst.
featured insights
I got a peek at our AI operations dashboard last week, care of Neal Iyer (thanks Neal). Like many of us, we are all trying to get a handle on the economics of AI. We already know that Tanya Brennan can spend $1,000 on Fable in one week without thinking about it. And we know that if we're not careful, our Amazon Web Services (AWS) Bedrock bill can double in a month. These cost components are the easiest to at least make visible. The total cost of AI, and the extent to which these costs are creating a return, continues to be a frustrating topic of conversation.
There is no one size fits all model for evaluating the cost of AI solutions and vendor pricing methodologies are changing. What is not changing (at least for the next 42 seconds) are the ten cost dimensions. I've grouped these into four buckets, and each dimension should be looked at for every major AI investment. A subset should be looked at for individual use cases, or small groups of use cases.
The most obvious cost elements are, of course, the application itself and the any use charges. We are pretty familiar with these as they are part of the traditional software-as-a-service (SaaS) model. This can be convenient pricing model with a good partner as it creates a predictable cost pattern. Whether you buy or build, you will likely have to integrate, which is your cost to deliver.
The choice of buy versus build has an impact on the downstream cost dimensions like talent. When you build, you have to get more of a very specialized type of talent that is difficult to find. Growing this talent internally is expensive and time consuming, partnering to acquire the talent is an option. Regardless, it is essential that every mortgage organization have some subset of W2 talent for AI and generative AI.
Oh the tokens. So much controversy, so much variability. A token is a word or part of a word, and it is the currency of large language models (LLMs). You pay for tokens as they go in, "thinking tokens", and the tokens that come out. Even if they are wrong.
The good news is that the cost of foundation AI has decreased significantly since "the early days" of 2023 and 2024. The price of a fixed level of AI capability has fallen 9x to 900x per year depending on the task (Epoch AI, 2025). GPT-4-level output that cost ~$30 per million tokens in 2023 now costs well under $1 (industry pricing analyses, 2026).
The bad news is that the cost of LLM usage for most organizations is still going up. Why? Because we typically use the newer bigger models for everything, without optimizing, and the price for those models has not gone down.
We recently implemented our AI ops dashboard for cost visibility, and have a range of techniques we use for cost optimization, including model distillation, and intelligent design of our innovation pipeline. Model distillation is the process of learning from a higher horsepower model and then training a lower cost model to act like the bigger one. Intelligent design in the innovation pipeline is the process of dynamically optimizing model choices and context management in the pipeline relative to the tasks they need to perform.
The next big challenge for us, and really for everyone, is how to connect token use to outcomes. We get the big bill from AWS, and we can connect that bill to people and to models, but we don't know if the spend was "worth it". This comes down to LLM observability and monitoring, and it's not easy but it can be done.
The first step is to find all the spend - subscriptions, cloud charges, vendor products, API charges - finding and gathering the cost for every person or team using generative AI. This sounds easy, but really isn't. For example, this may mean combing through credit card statements for that pesky Synthesia subscription you forgot you had. And then you have to do this across all credit cards.
This step requires you to connect the dollars spent to the people who spent them. You found that $2,500 charge for Replit in May - now you have to connect that to the people who have access. You have to find the individuals who are using the product compared to ilde seats and users. This requires you to go into yet another platform, and connect that platform to you AI operations solutions. This can make your team nervous. People generally do not like to be monitored. And the point is not to monitor the people, it is to understand how AI is being used for the purpose of optimizing cost. You may even need some teams or people to spend MORE.
This is the hardest part, and requires an even lower level of monitoring as you want to see not only THAT the people are using AI but WHAT they are using it to create. If you only use one platform, say Microsoft Copilot, this won't be that hard. But if you use a lot of different platforms, this can be a challenge. We have recently implemented Langfuse for LLM observability and have started to connect prompts to tokens, models, tool use, and agent invocation.
So now you found all the spend, connected it to your AI ops console, and are able to observe the relationship between the dollars and the purpose. Next, and most difficult, you elevate that to the outcome level. Full transparency, we are still working on this. We have pockets of use that we understand well, and pockets that we still have to connect. Spend + person + intent = the basis for understanding the outcome.
This is your data, your evaluation process, and your compliance approach. Much has been said about the important of good data, I won't dwell on this here. But I will talk about the other costs of quality - compliance and evaluation. Cost of compliance is the cost of the program and processes you have in place to govern your AI use - and especially your high risk uses.
Step one is to understand what you are doing with AI and does any of it qualify as what Colorado is calling a consequential decision. I use this standard because it's the best one I can find, and it works well with what has been defined by the Administration for compliance by the federal government. The decision to place something in the catalog as well as the decision to designate a use case as consequential or high risk should be documented.
I am advising everyone to put in place the governance model proposed by the Administration in M-25-21, I think it's a solid framework that is straightforward. The hardest elements are the logging and monitoring, both of which needs to be in place for any AI use case in my opinion.
One thing to note, a human-in-the-loop who just reviews AI generated content will ultimately be insufficient to meet what I think examiners and regulators will look for. In addition, you will want to understand, document, and be able to explain and demonstrate effectiveness of your context specific guardrails.
In addition to guardrails, another key cost of compliance is your evaluation program. These are the tests, preferably performed at scale by automation, that you perform to test that AI systems are performing in accordance with your expectations - including standards for quality. If you do not have an evaluation program, start small, with core metrics that matter.
There really is not substitute for high quality benchmark data sets. These are sometimes also called QA (question and answer) pairs. QA pairs are are structured data sets documenting representative test questions along with their verified, correct reference answer (often called "ground truth"). They are used as a standard benchmark to test how accurately and safely an AI model or retrieval system responds. These are time consuming to do well but will be a critical part of your ongoing quality process.
You want to run your evals on a schedule, anytime you need to test prompt changes, and then before and after model upgrades. Depending on what information your vendor makes available to you, you might be able to rely on their evaluation program but at the end of the day, you are accountable. As always, I recommend the open source framework by Promptfoo
As you run your evals, you might see that your eval scores are getting worse, you will need to take action if this is the case. Your eval program will be a critical investment in the robustness of your overall AI ecosystem and should not be skimped on. It may end up being a significant cost to consider.
And finally, the cost of innovation. This article is already way too long so I'll summarize. Talent, time, change, and opportunity cost. These are the hardest to quantify and the most important to do well. I promise to write another article on this specifically as part of this series.
featured insights
ARLINGTON, VA, UNITED STATES, June 9, 2026 -- PhoenixTeam announced the launch of Human Futures, a new initiative helping young people prepare for a world being reshaped by artificial intelligence. Human Futures is grounded in a single belief: the next generation will inherit a different world than the one we grew up in, and it is our responsibility to prepare and employ them. Through AI education, career readiness, employer-connected learning, and hands-on project work, the initiative helps students and early-career professionals develop the skills, confidence, and judgment to use AI thoughtfully, solve real problems, and shape what comes next.
For PhoenixTeam, Human Futures is a natural extension of the company’s work helping organizations understand and apply AI in practical, responsible ways. After years of building AI-enabled solutions, educating mortgage and technology professionals, and helping clients navigate generative AI adoption, PhoenixTeam is now applying that experience to a broader challenge: helping people prepare for the world AI is creating.
The first Human Futures program is underway through PhoenixTeam’s 2026 Summer Internship Program. The cohort includes interns working across Human Futures, AI services and Phoenix Burst, project delivery, business operations, and marketing. Throughout the program, interns receive weekly AI education, learn how businesses operate, build communication and career-readiness skills, and work on individual and team projects that will become part of a real portfolio.
“All of the content we deliver in our Human Futures programming will be accessible to all and available for free. We hope this will be the first of many ways that we can use fear to catalyze action,” said Tela Gallagher Mathias, CTO of PhoenixTeam and CEO of Phoenix Burst. “Human Futures is about bringing the focus back to people. We want young people to understand AI, work with it responsibly, and still build the human skills that make them valuable: courage, judgment, aspiration, resilience, and authentic human connection.”
“The future of work is changing quickly, but preparation cannot just be about tools,” said Tanya Brennan, CEO of PhoenixTeam. “Human Futures is about helping people build the thinking, confidence, and real-world experience they need to participate in that future. The first cohort is already showing what happens when young people are given real problems, real mentors, and room to build.”
Human Futures will continue to grow through programs for students, early-career professionals, educators, families, and employers. The initiative will focus on practical AI readiness, responsible use, portfolio-based learning, and employer-connected pathways that help participants move from learning to contribution.
To learn more, visit www.aifutures.com.
About Human Futures
Human Futures is a PhoenixTeam initiative preparing learners to think, work, and lead in a world being reshaped by artificial intelligence. Built for learners, parents, educators, and employers, Human Futures connects AI education with real-world projects, mentorship, career readiness, and portfolio-based learning. The initiative is rooted in a human-first belief: AI should expand what people are capable of, not replace the thinking, creativity, judgment, and resilience that make people valuable. Through hands-on programs and employer-connected pathways, Human Futures helps learners build practical skills, confidence, and real experience for the future of work and life. For more information, please visit www.aifutures.com
About PhoenixTeam
PhoenixTeam is a woman-owned technology services firm headquartered in Arlington, Virginia, specializing in AI-powered mortgage operations and technology services for the mortgage and financial services industries and federal housing agencies. Our mission is to enable affordable and accessible homeownership through innovative, customer-centric technology. With a strong focus on generative AI, we tackle complex industry challenges, equipping businesses with cutting-edge tools that enhance innovation, efficiency, and compliance. By bridging the gap between technology and business teams, we strive to bring joy and purpose back to software development, making a meaningful impact in the lives of our clients and homeowners everywhere. For more information, please visit www.phoenixoutcomes.com.
