About SurfaceAI
SurfaceAI is the first AI agent platform for multifamily operations, helping property management teams boost efficiency, uncover hidden revenue, and make faster, smarter decisions.
Context
Vertical AI represents a paradigm shift in enterprise software—one poised to eclipse even the massive $300 billion market created by vertical SaaS. Unlike previous waves of industry-specific software, AI applications can tackle high-cost, repetitive, language-based tasks that dominate mature, legacy industries.
This case study explores how Stackpoint methodically transformed a hypothesis about AI's potential in property management into SurfaceAI, an AI agent platform that secured significant seed funding largely based on the strength of systematic discovery work.
The Opportunity: AI in Property Management
When identifying industries ripe for AI transformation, Stackpoint analyzes specific characteristics that indicate potential for disruption. Property management checked all the boxes:
Significant data fragmentation across systems
High volume of repetitive tasks
Substantial financial impact when things go wrong
Starting with the Right Founder and Team
SurfaceAI began with a hypothesis: there is a huge opportunity in the market, but the exact solution wasn’t yet defined. Recognizing that building a sophisticated AI platform required exceptional technical leadership, Stackpoint recruited Jason Wallis, former CTO and Founder of Mozu (acquired) and veteran of senior engineering roles at Amazon, Dell, and Wayfair to be the CEO of SurfaceAI. From day one, Jason and the Stackpoint team worked side by side, shaping the vision and building SurfaceAI from the ground up.
"From day one, I had access to everything needed to move quickly—user research, product development, design expertise, back-office infrastructure for company formation, and marketing support. Once the studio team was assigned, we immediately engaged in an intensive kickoff phase to align our vision and strategy. From the moment we said go, we were all attached at the hip. And it felt like just one functioning unit.” — Jason Wallis
Different from other venture studios, Stackpoint provides an elite cross-functional team of product managers, designers, engineers, and operators to develop the initial hypothesis and leverage Stackpoint's extensive network to connect with property owners and operators to validate and test the minimal viable product.
Unlocking Deep Problem Discovery Through Strategic Access
For most early-stage founders, securing meaningful customer access for deep research is a major challenge. Understanding complex operational environments requires more than surface-level conversations—it demands hands-on immersion.
Stackpoint’s industry network provided a crucial edge, granting Jason and the team rare access to owner/operators managing over 200,000 units, tapping into real-world data and subject matter expertise that would be nearly impossible for a solo founder to access. They were also able to go onsite to properties where they shadowed property managers and maintenance staff throughout their workday to gain firsthand insights.
The team didn’t just observe, they worked alongside the onsite team, inspecting units and tackling tasks firsthand. “We were doing the job with them to build trust and truly understand their pain points,” says Cristina Jain, Stackpoint VP of Product. This approach surfaced two major inefficiencies:
Workflow bottlenecks – Property teams struggled with task prioritization, constantly juggling competing demands without clear guidance on what created the most value.
Fragmented information – Critical data was scattered across multiple systems, forcing staff to waste hours searching instead of acting, leading to missed opportunities and delays.
After compiling their findings, the team presented them to property management executives. “Every single executive was blown away,” Jason recalls. The response validated that these operational pain points weren’t just frustrating—they were materially impacting efficiency and financial performance.
“In just weeks, Stackpoint got me in front of some of the most strategic customer prospects in the industry—people who are normally impossible to reach. That early access didn’t just shape the product, it accelerated our timeline by years and paved the way for our first few million in revenue.” — Jason Wallis
Rapid Prototyping: Testing Without Coding
The Stackpoint team took a unique approach to rapid prototyping that didn't require writing code. They embraced extreme low-fidelity prototyping to gather actionable insights efficiently.
Instead of wireframes or mockups, the team created functional paper prototypes that used real customer data to simulate the eventual product experience. Each night, they would manually pull data from property management systems, organize it according to their envisioned product structure, and provide printed versions to property teams the following morning.
"That's how we built conviction on whether the concept was valuable, because we were able to see them engage with it on paper before building it on the screens," explains Cristina. "They were carrying it all day with them and taking notes on it."
The property managers' enthusiastic adoption spoke volumes. Staff integrated these paper tools into their daily workflows immediately, annotating them with additional information and using them to prioritize their activities. This visible behavior change provided compelling evidence that the concept addressed real needs.
"Stackpoint has a rigorous process on problem discovery and rapid prototyping. What we accomplished with these paper prototypes was remarkable—they validated our concept without writing a single line of code. This shortened our path to product-market fit so we can get to commercial traction faster than we normally could” explained Jason.
Beyond validating the core concept, this approach revealed specific user preferences that would have been difficult to anticipate:
Which metrics users looked at first
How they prioritized different types of information
What additional context they needed for decision-making
Where they added notes and supplementary data
This paper-first methodology allowed the team to avoid expensive development detours while building a detailed blueprint for the eventual product—all while strengthening relationships with their design partners through daily interaction and immediate value delivery.

Refining the Solution Through User Feedback
With the paper prototype validated, the team moved to content organization sessions with users. They used whiteboards and sticky notes to determine what information should appear on each screen, in what order, and why.
"We had all of the users at multiple different sites and properties talk us through what they wanted to see, why, when information was important, what wasn't, to really inform the architecture," explains the team.
This user-centered approach continued throughout development, with low-fidelity prototypes tested with real users. Even when they built a simple Q&A function with limited data, users asked numerous questions, revealing how they would actually use the system in practice.

Building the AI Architecture
Stackpoint's technical team focused on three key challenges:
Data ingestion and normalization: Property management companies use different systems that don't naturally communicate. The team created a normalization layer to make disparate data usable.
RAG pipeline: The team built a system to determine what type of question was being asked and accordingly choose the right technical process to get the correct answer.
Prompt engineering: The team optimized prompts to minimize hallucinations and format responses appropriately for different use cases.
To facilitate this work, Stackpoint built internal tools that broke down the AI process into discrete steps, allowing them to adjust prompts at each stage and improve response quality iteratively.
"We built an internal tool where you could actually see the coding in each of these steps, and you could adjust the prompts in each of the steps live and then rerun it to see if you were getting a better result," explains Cristina.
From Discovery to Funding: The Stackpoint Difference
The disciplined discovery process yielded impressive results. When it came time to pitch investors, the team had a compelling story backed by substantial evidence:
Deep understanding of user problems and workflows
Documented user validation through multiple testing phases
Specific quotes and feedback from design partners
Clear technical approach to solving the identified problems
This foundation was so compelling that Bessemer Venture Partners preemptively backed the company at the seed stage—before the product was even built.
"As we went to investors, we weren't pitching a product—we were presenting a body of evidence. The deck was filled with real user insights, validation, and a clear case for why this problem needed to be solved, right now, with AI," explains Jason. “Stackpoint's disciplined approach to discovery gave us credibility that most early-stage startups simply don't have.”
The Journey Continues: From MVP to AI Agent Platform
The methodical discovery process laid the groundwork for what has become an expansive product vision. SurfaceAI evolved from addressing specific pain points to creating an integrated AI platform for property management. The initial insights uncovered during those early property visits directly shaped what is now SurfaceAI's Intelligent Workspace - a command center for multifamily operators.
Through Stackpoint's systematic approach, SurfaceAI:
Engaged with owner/operators of 200,000+ units to validate hypotheses
Conducted 25+ on-site sessions with property management teams
Signed their first contract in less than 15 months after launch
The Intelligent Workspace serves as the central interface where users interact with the system - asking questions, receiving insights, and managing priorities without switching between multiple systems. Complementing this, SurfaceAI's specialized AI Agents work in the background, handling specific high-value workflows like lease audits, payment collection, and renewals.
For example, the Lease Audit AI Agent continuously monitors leases to catch revenue leaks before they happen, automatically identifying discrepancies in rent amounts, missing amenity charges, or unapproved discounts. Meanwhile, the Delinquency AI Agent monitors overdue payments and handles follow-up communications, freeing property managers from the time-consuming process of payment collection.
Methodical Product Discovery: The Key to Vertical AI Startup Success
The SurfaceAI story exemplifies Stackpoint's firm belief that systematic discovery and validation in collaboration with design partners is crucial for building successful ventures in mature, legacy industries. By deeply understanding specific industry challenges, validating solutions through creative low-cost methods, and building authentic relationships with domain experts, Stackpoint builds ventures from scratch that deliver true value.
This methodical approach ensures companies are moving in the right direction before accelerating, positioning them in the top right quadrant:

The discovery-first philosophy is particularly powerful for vertical AI applications where domain expertise, user trust, and workflow integration often present greater challenges than the technical implementation itself. As AI continues to transform traditional industries, those who can most effectively apply AI to solve well-defined, high-value problems for specific users will win.
SurfaceAI's journey from concept to funded company demonstrates how Stackpoint's thorough problem exploration creates the foundation for technology that truly matters—and how their cross-functional team approach helps founders navigate the complex path from idea to successful company.
"What Stackpoint offered was amazing—a comprehensive, proven system for company building," reflects Jason. "From day one, I had access to a complete team with specialized expertise across product, design, engineering, and go-to-market. In my career, I've rarely encountered an organization with such a concentration of exceptional talent. Working with the Stackpoint team, I'm surrounded by high-integrity, highly capable professionals who are deeply invested in SurfaceAI's success and ready to help at every turn." — Jason Wallis