Dentistry
The second vertical expansion after adults. Bigger market by practitioner count, different pain pattern, same one-pane-of-glass thesis.
Why dentistry is a real market
| Segment | US practitioner count | Average practice size |
|---|---|---|
| Dentists | ~200,000 | Small, often solo or 2–4 chair |
| Pediatric dentists | ~7,500 (subset) | Solo or small group |
| Orthodontists, oral surgeons | ~10,000 each | Specialty practices |
The dental market is bigger than primary-care medicine on practitioner count alone, and dental practices are typically smaller and more fragmented than medical practices — which is precisely the segment where vertical-purpose-built software wins against generic EHR/PMS suites.
Why the pain is real
The administrative burden in dentistry is arguably worse than in primary care:
- Insurance complexity. Dental insurance has its own ecosystem (PPO networks, fee schedules, pre-authorizations) that's even more fragmented than medical. A typical dental practice deals with 20+ payer plans concurrently.
- Patient scheduling chaos. Recall management (every 6 months for hygiene), no-show rates, hygienist-vs-dentist scheduling complexity, multi-operatory chair-time optimization.
- Lab coordination. Crowns, dentures, orthodontic appliances all flow through external labs with their own communication protocols, turnaround SLAs, and cost-tracking. Most dental software handles this poorly.
- Imaging integration. Intraoral cameras, panoramic X-rays, cone-beam CT — image sources need to land in the chart, version, annotate, and survive payer audits.
Existing dental practice management software (Dentrix, Eaglesoft, Open Dental, Curve) is dominated by Henry Schein and a handful of legacy players, all built before AI was a substrate.
What carries over from medical Starlight
- Patient + Family billing model — applies cleanly. Family dental plans, multi-child orthodontics billing.
- Subscription / membership pricing — dental DPC equivalents are growing (in-house dental membership plans), and our family/household billing model is a natural fit.
- AI scribe (with adaptation) — dental clinical notes are shorter and templated, but Claude can absolutely draft them. Bigger leverage may be in post-visit treatment-plan summaries for patients (analog of our parent-app summary).
- Single-pane-of-glass thesis — same wedge.
- Document auto-routing — dental practices receive insurance EOBs, lab cases, referral letters — same Claude name+DOB-extraction pattern works.
- Compliance posture — HIPAA applies to dental practices identically; SOC 2 carries over.
What's different
- Imaging is a much bigger surface. Dental software lives or dies on intraoral imaging integration (Dexis, Carestream, Sopro). We'd need to integrate at least one major DICOM-equivalent dental imaging vendor.
- Lab case management — a workflow that doesn't exist in medical Starlight at all. Lab orders, lab cases in transit, lab returns, lab cost reconciliation — all needs to be a first-class entity.
- Treatment planning is the central artifact, not the SOAP note. Dental visits are organized around multi-visit treatment plans: "$3,400 of work over 4 visits, sequenced this way." Patient-facing treatment plan presentation and acceptance flow is a marquee feature.
- Hygiene recall is the steady-state economic engine of a dental practice. Recall-management workflows + automated patient-app prompts are essential.
- Less tech-savvy buyer. General dentists tend to be more conservative on software adoption than DPC primary-care docs. Sales motion is harder. Counter: when one of our pediatric DPC docs has a dentist spouse or referral relationship, that's an easier introduction than cold sales.
- Schedule II isn't really a thing in dentistry, but DEA + state-specific opioid monitoring still applies (oral surgery, post-extraction Rx).
Pediatric dentistry as the wedge
Pediatric dentistry is to general dentistry what pediatric DPC is to general primary care: harder UX (parent + child), more emotionally engaged buyers, smaller and more fragmented practices, lower current software quality. It is the natural beachhead for porting Starlight to dentistry, because the parent-app + family-billing + Guardian-as-first-class-entity work we did for pediatric DPC ports directly.
If Erik wants to take this seriously, the right pilot is a pediatric dentist — possibly someone Dr. P can introduce us to in the Austin pediatric care community.
Competitive landscape (sketch)
| Incumbent | Strengths | Weaknesses |
|---|---|---|
| Dentrix (Henry Schein) | Dominant market share, deep imaging integration | Legacy desktop, on-premise dependence, weak AI |
| Eaglesoft (Patterson) | Bundled with Patterson supply contracts | Same legacy issues |
| Open Dental | Open-source, popular with smaller practices | Self-host complexity, weak patient-facing surface |
| Curve Dental | Cloud-native (one of the few) | Smaller scale, limited AI |
| Dental Intelligence, Modento, Weave | Patient-engagement add-ons that bolt onto the above | They're add-ons. We are the platform. |
The wedge against dentistry incumbents is the same as against Atlas.MD: AI substrate from day one, native patient app, single pane of glass, family billing.
Open questions
- How dental-specific are the imaging integration costs? Likely high enough that the first dental customer needs to fund or co-engineer the imaging connector.
- Lab case management is a net-new entity we'd have to design. Is there a clean abstraction that also serves medical referrals (which we do already need)?
- Do we go pediatric-dentistry first (cleanest port from peds DPC) or general-dentistry-first (bigger market, harder sales)?
This bet is gated behind adults DPC working. Pediatric DPC → adult DPC → pediatric dentistry → general dentistry is the pattern that maximally reuses each prior step's product surface area.