AI Readiness Assessment for a West-Coast Dental Group Modernising Intake and Operations
Let's Connect
Overview
What we built
A 42-office dental group had standardised its systems but not its AI ambitions, with offices trialling scheduling and billing tools ad hoc and patient data quality varying widely. A structured assessment gave leadership one scored roadmap they could actually afford.
In plain terms: this dental group runs 42 offices on the US West Coast, and had already done the hard infrastructure work of standardising on one shared practice-management system and digital patient intake. What it had not done was agree on AI. Individual offices were trialling scheduling and billing tools on their own, patient data quality varied widely from location to location, and leadership had no way to judge which AI investments the group could realistically absorb.
We ran a 10-week AI readiness assessment across the group: auditing data quality and integration points in the practice-management and intake platforms, interviewing office managers and clinical leads at 15 representative offices, and scoring the organisation against a six-dimension readiness framework. The result was a sequenced 18-month roadmap of AI initiatives, each with a budget range and prerequisite data work attached. From 31 candidate use cases, 7 were approved for the first-year roadmap, and the first initiative, automated recall scheduling, moved to a funded pilot within 8 weeks of the final readout.
The Problem
Scattered pilots, no AI foundation
The group had done the unglamorous work first, standardising 42 offices onto a shared practice-management system and digital patient intake. That should have made AI easier to plan for, but instead each office had started experimenting on its own, trialling scheduling and billing tools ad hoc with no shared method for judging whether any of them worked.
Underneath the shared systems, patient data quality varied widely between locations. Some offices kept clean, complete intake records; others did not, and nobody had measured the gap between them. That mattered because any AI initiative would need to trust the data it was built on, and the group had no evidence either way.
Leadership faced a familiar problem without familiar tools to solve it: no way to judge which AI investments the group could realistically absorb, given 42 offices at different levels of readiness, ad hoc pilots already under way, and no framework for weighing one opportunity against another.
Ad hoc office trials
Individual offices were trialling scheduling and billing tools on their own, with no shared method for comparing results across the group's 42 offices.
Uneven data quality
Patient data quality varied widely between locations, and nobody had measured the gap, leaving any future AI build resting on unverified foundations.
No readiness framework
Leadership had no structured way to judge which AI investments the group could realistically absorb, so decisions defaulted to instinct rather than evidence.
Standardised systems, no plan
The shared practice-management system and digital patient intake gave the group a technical foundation, but no plan existed for what to build on top of it.
What it was costing them
Every office running its own ad hoc trial was time and budget spent without a shared method for judging the result, and every office with incomplete intake data was a risk waiting to surface the moment an AI initiative depended on it. With 42 offices at different levels of readiness and no framework to compare them, the group risked funding the loudest pitch rather than the one most likely to work.
The Solution
Structured readiness assessment
We ran a 10-week AI readiness assessment designed to answer two questions at once: what is the group's data actually capable of, and which AI initiatives are worth funding first. The first step audited data quality and integration points in the practice-management and intake platforms, giving the group its first real measurement of where records were solid and where they were not.
We then interviewed office managers and clinical leads at 15 representative offices, capturing the practical, day-to-day view that a systems audit alone could not surface, and scored the organisation against a six-dimension readiness framework covering data, process and people together rather than technology in isolation.
The assessment produced a sequenced 18-month roadmap of AI initiatives, each carrying a budget range and the prerequisite data work it would need before it could start. Of 31 candidate use cases identified, 7 were approved for the first-year roadmap, and we helped move the first, automated recall scheduling, into a funded pilot within 8 weeks of the final readout.
Key decisions
Audit data before scoring ideas
Data quality and integration points across the practice-management and intake platforms were measured first, so later scoring rested on evidence rather than assumption.
Talk to 15 representative offices
Interviews with office managers and clinical leads at 15 offices grounded the assessment in day-to-day reality, not just system-level data.
Score on six dimensions
The six-dimension readiness framework judged data, process and people together, avoiding a roadmap built on technology readiness alone.
Attach budgets and prerequisites
Every one of the 18-month roadmap's initiatives carries a budget range and the prerequisite data work it needs, so approval means something concrete.
Fund the first pilot immediately
Automated recall scheduling, the first roadmap initiative, was moved into a funded pilot within 8 weeks of the final readout rather than left to wait.
Measurable Impact
What changed after launch
The assessment gave the group a roadmap sized to what it could actually absorb. Of 31 candidate AI use cases identified across the 42 offices, 7 were approved for the first-year roadmap, each with a budget range and prerequisite data work attached rather than an open-ended ambition.
The data audit also surfaced a problem the group could now act on: intake records were fully complete for only 64% of active patients, which triggered a remediation programme before any AI build could rely on that data. Consolidating 3 duplicate ad-hoc AI pilot subscriptions cut related software spend by 22%, and the first roadmap initiative, automated recall scheduling, reached a funded pilot within 8 weeks of the final readout.
AI initiatives
Ad hoc trials at individual offices, no shared plan
31 use cases scored, 7 approved for the roadmap
Data quality
Varied widely between locations, never measured
Measured at 64% complete, remediation programme funded
Tooling spend
3 duplicate ad-hoc AI pilot subscriptions running
Consolidated, cutting related software spend by 22%
Pilot delivery
No AI initiative had a funded starting point
Recall scheduling pilot funded within 8 weeks
Headline results
31 candidate AI use cases catalogued and scored, with 7 approved for the first-year roadmap
Data audit found intake records fully complete for only 64% of active patients, triggering a remediation programme before any AI build
3 duplicate ad-hoc AI pilot subscriptions consolidated, cutting related software spend by 22%
First roadmap initiative, automated recall scheduling, moved to a funded pilot within 8 weeks of the final readout
Tech & Tools Used
What powered the build
Every tool below earned its place in this engagement. Here is the part each one played.
Python (pandas)
Profiled intake and scheduling records during the data audit, quantifying the gap in completeness across the group's 42 offices.
SQL Server
Held the practice-management system's records that the audit examined for data quality and integration readiness.
PostgreSQL
Stored consolidated extracts from the intake platform, giving the audit one place to measure completeness across offices.
Airbyte
Replicated data out of the practice-management and intake platforms without custom pipeline work, so the audit could start quickly.
dbt
Modelled the readiness scoring logic behind the six-dimension framework, keeping the scores consistent across all 42 offices.
Metabase
Gave office managers and clinical leads self-serve views of the audit findings during the 15 representative-office interviews.
Jupyter
Used to analyse patterns in the 31 candidate use cases and test which were feasible given the measured data quality.
Miro
Hosted the scoring sessions where the 31 candidate use cases were ranked against the six-dimension readiness framework.
Jira
Carries the 18-month roadmap's initiatives and the recall scheduling pilot's backlog, with budget ranges and prerequisites attached.
Ready to Build your Dental Healthcare Business with AI Readiness Assessment
Ask Byte
Ask Byte
Typically replies instantly
just Now
Hi! I'm OrganByte's assistant. How can I help you today?
AI-generated content may be incorrect
