← All case studies Healthcare · Case study

An AI-Native Dental PMS Built in 10 Hours

How an AI-native dental practice management software platform got 75–90% built in about 10 hours for roughly $500 in AI tokens — a full PMS with patient management, scheduling, revenue cycle management, and X12 claims processing, engineered to compete with Dentrix, Eaglesoft, and Open Dental.

Client
Oscar / Canopy Dental
Industry
Healthcare
Timeline
8–16 weeks
Headline result
~10 hrs
~10 hrs
Build Time
~$500
Total Token Spend
~7,000
Lines of Backend Code
3
Live API Integrations
About the Company

An AI-native challenger building cloud dental PMS software to compete with Dentrix, Eaglesoft, and Open Dental.

Oscar / Canopy Dental is an AI-native dental software company building a next-generation practice management system for dental practices and DSOs. CEO John Salter and the founding team set out to compete directly with the legacy PMS platforms that dominate the market: Dentrix, Eaglesoft, and Open Dental, which together run on roughly 55% of dental practices in North America. The bet is on an AI-powered PMS that's cloud-native from day one, with a conversational AI interface in place of the menu-driven workflows those legacy systems still rely on.

The Problem

Building a full dental PMS from scratch at the pace and cost a new entrant can actually survive.

Entering the dental PMS market is expensive. A modern system has to handle patient management, scheduling, dental RCM, real-time insurance eligibility verification, X12 electronic claims processing, multi-tenant architecture for DSO deployments, and integrations across clearinghouse and billing APIs. Built traditionally, that's a 9 to 18 month timeline and several hundred thousand dollars before a single practice logs in.

Oscar's original engagement reflected that math: a $250K contract structured over 9 months, with an internal technical stakeholder embedded in the build. For a new entrant going up against Dentrix and Eaglesoft, both backed by billion-dollar supply distributors, that timeline was the constraint. Every month of build was a month not competing. And every dollar spent on scaffolding was a dollar not spent on the AI work that was supposed to win the market.

The Solution

An AI-native dental PMS, built at a pace legacy vendors can't match.

CustomAI Studio took ownership of the architecture and AI implementation and delivered a working prototype of the full platform: patient records, appointment scheduling, dental RCM, real-time eligibility verification, X12 claims processing, multi-tenant auth, and the conversational AI chat interface. CAIS Head of Product Allan Crawford built 75–90% of the system over roughly 10 hours of AI-accelerated development, at a token spend of about $400 to $500.

The backend is around 7,000 lines of code. Authentication runs through Clerk with a multi-tenant model: a master database for auth and isolated firm-specific data volumes, built specifically for DSO deployments where multiple locations need centralized control with partitioned data. The insurance layer pairs Steady and Vine, since Steady alone doesn't support x-ray attachments and no dental claims platform ships without them. Next Health handles bidirectional sync via its beta API. All three integrations were functional in sandbox at handoff.

The AI chat interface is the MVP differentiator. Practice staff interact with the PMS conversationally instead of working through the dense menu trees that define Dentrix, Eaglesoft, and Open Dental. Deployment runs on GCP for MVP, with an AWS migration planned for enterprise and DSO rollouts. Competing with legacy dental PMS isn't about matching feature lists. It's about making the legacy feature list feel dated, and that's what the architecture is built to do.

The working prototype shifted the engagement. After CustomAI Studio demonstrated the full system (patient management, RCM, eligibility, X12 claims, multi-tenant auth, and conversational AI, all functional in sandbox) the original $250K / 9-month contract was restructured into a roughly 4-month partnership. Oscar closed and wired payment in late February 2026.

The Impact

A full AI-powered dental PMS, built and integrated in a fraction of the time and cost the category has typically demanded.

The headline numbers say most of it: 75–90% of a production-grade dental PMS, built in about 10 hours, for $400 to $500 in AI tokens. Around 7,000 lines of backend code. Three live API integrations. A working prototype of a system that, built traditionally, would have consumed the better part of a year and a quarter-million dollars before reaching the same state.

For Oscar, the impact extended past the build. The original $250K / 9-month engagement was restructured into a roughly 4-month partnership after the working prototype was delivered, which collapsed time-to-market and freed capital that would otherwise have been locked in scaffolding work. The roadmap now points at what actually differentiates an AI-native PMS from Dentrix, Eaglesoft, and Open Dental: the conversational AI layer, the multi-tenant DSO architecture, and the bidirectional sync that lets practices migrate without ripping out imaging or billing infrastructure on day one.

The dental PMS market has been defined for three decades by on-prem incumbents with deep moats. The pace at which a credible challenger can now ship a full system changes what's possible for new entrants.

Under the hood.

The system architecture behind the Canopy / Oscar Dental engagement — how data flows across the practice's tooling and into the AI operating layer that automates day-to-day operations.

Architecture · Canopy on AgenticOS Canopy Dental on the agentic OS — runtime architecture and data migration in one view. Canopy Dental Rev 01 · 2026.04 Author Oscar · Syncronizer A Runtime architecture Live system 01 — PLATFORM Canopy Dental Practice management surface 02 — AGENT Oscar Agent runtime SATELLITE Tool registry Dynamic tools 03 — DATA · CRITICAL PATH agentic Database Single source of truth · shared across runtime and migration SHARED ↔ B 04 — INPUTS Data Sources Practice records Integrations Third-party systems B Data migration Syncronizer · One-time + ongoing 01 — DESTINATION · CRITICAL PATH agentic Database Same instance — receives migrated records SHARED ↔ A 02 — ENGINE · SYNCRONIZER Data Migration Normalizes · dedupes · writes to agentic Database 03 — SOURCES Data Source Legacy PMS Data Source Imaging · charts Data Source Billing · claims Critical hinge / agentic Database is shared across runtime and migration 2 views · 1 shared store · runtime + migration

Results.

  • ~10 hrs — Build time for 75–90% of the system
  • ~$500 — Total AI token spend
  • ~7,000 — Lines of backend code
  • 3 — Live API integrations (Next Health, Steady, Vine)
  • 55% — Share of dental PMS market held by legacy on-prem systems being targeted

Ready to become
AI-Native?

Book a 30-minute conversation. We'll map the highest-leverage workflows in your business and tell you whether AI is the right answer.