iOS Project
MakenaAI Appraiser Platform
Enterprise iOS platform for real estate appraisers. Subscription and payment system that onboarded 2,000+ clients within one week of launch. Led the "Digital Clone" AR pilot for remote property inspections using ARKit — Sean's first role as sole software engineer.
The context
MakenaAI is an AI, mobile, and property tech company serving real estate appraisers. Sean joined as the sole software engineer on a suite of high-impact AI, mobile, and AR projects — working directly with non-technical appraiser stakeholders and presenting adoption roadmaps to leadership.
The role spanned the full product lifecycle: technical discovery with stakeholders, architecture and implementation, and go-to-market support. Three distinct product tracks shipped during this tenure.
Premium subscription system
Architected and shipped the premium subscription and payment system for the MakenaAI appraiser iOS application. The system used StoreKit for in-app purchase management, with server-side receipt validation against the FastAPI backend.
The key delivery metric: 2,000+ appraiser clients onboarded within one week of launch. This wasn't a soft launch — it was a coordinated release to an existing customer base that was already familiar with the product and waiting for the subscription tier. The system had to work correctly and reliably from day one.
Digital Clone AR pilot
Led the "Digital Clone" AR pilot — an enterprise product for remote property inspections that eliminated the need for in-person appraisal visits in applicable scenarios. The concept: appraisers could capture a property remotely using ARKit's world tracking and plane detection, creating a structured spatial model of the property that could be reviewed and measured from any location.
The translation challenge: real estate appraisers are not technical users. They know property inspection workflows intimately but have no background in spatial computing. Engineering the experience to feel like an extension of their existing workflow — not a new paradigm they had to learn — was the core UX challenge.
ARKit world tracking
Persistent spatial anchors for room boundaries, measurement points, and structural features. Accurate spatial capture without specialized hardware.
RealityKit rendering
Real-time 3D visualization of the captured space, with measurement overlays and annotation tools accessible to non-technical users.
AI model routing
Translated complex end-user appraiser workflows into a model routing architecture — AI assistance for measurement interpretation and feature identification, with appraiser review and approval gates.
Team capability training
Led AI capability training for the engineering team — establishing patterns for integrating AI assistance into the AR workflow without creating over-reliance on model output for legally-significant appraisal decisions.
Internal RAG chatbot
Beyond the mobile work, Sean also built MakenaAI's internal RAG chatbot from requirements through production — a natural language interface for the appraiser knowledge base. Led technical discovery with non-technical stakeholder appraisers, built the system, and presented the adoption roadmap to leadership.
The appraisal domain has a significant knowledge density: regulatory requirements, methodology guidelines, local market conditions, and property classification rules vary by jurisdiction and change regularly. The RAG architecture made this knowledge accessible through natural language, eliminating lookup friction for appraisers during active valuations.