Two EHRs, One Operating Picture: an AI Readiness Assessment for a Behavioral-Health Practice
An independent psychiatry practice running two EHRs wanted to know where AI actually fit. We mapped the workflows, ranked the opportunities, and told them which automations weren't worth building.
Executive summary.
An independent behavioral-health practice in Richmond, VA — owner-operated, with a small clinical and admin team — brought us in for an AI readiness assessment in spring 2026. Not a build. A look.
Before we ever showed up, the practice completed a 38-question structured discovery questionnaire across 10 sections, delivered as a PIN-gated web app that autosaved and fed structured data straight into our CRM. No PDF email round-trips, no “did you get my attachment?” From there we mapped workflows across their dual-EHR setup, built an AI opportunity map ranked by impact, complexity, and compliance risk, audited their marketing visibility, and ran two independent research passes on an EHR consolidation decision.
What they got: a clear operating picture and a prioritized roadmap — including an honest list of automations we told them not to pursue, because their closed, no-API systems made those builds infeasible. We’d rather flag that on page one than discover it three invoices in.
The challenge.
The practice runs two different EHRs for different service lines. That’s not an exotic problem — it’s a common one — but it has a daily cost: staff manually re-keyed patient demographics between the two systems. For one admin role, that re-keying alone ate roughly 5 hours a week.
Beyond the duplicate data entry, the owner had the question every small practice has in 2026: where does AI actually help here, and where is it a distraction? In healthcare that question comes with teeth — HIPAA isn’t a footnote, and a tool that’s clever but non-compliant is worse than no tool at all.
There was also a visibility problem nobody had quantified. When we audited the practice’s marketing footprint, the site scored 3.2/10 on SEO and 2.4/10 on AI-assistant visibility — meaning when prospective patients ask a search engine or a chatbot for care in the area, this practice barely registers. We mapped 16 local competitors to show exactly who does.
The Ironworks approach.
Structured Discovery
A 38-question questionnaire across 10 sections, delivered as a PIN-gated web app. Autosaving, completable in pieces, structured responses flowing directly into our system — so the kickoff conversation started from facts, not guesses.
Workflow Mapping
We traced how work actually moves through the dual-EHR setup — where data gets entered twice, where staff bridge systems by hand, and which of those gaps software can realistically close.
Opportunity Map and Visibility Audit
Every candidate AI use case ranked by impact, complexity, and compliance risk — HIPAA first. Where closed, no-API systems made an automation infeasible, we said so instead of pitching an impossible build. In parallel: the SEO and AI-visibility audit with the 16-competitor map.
Delivery and Roadmap
Deliverables presented as interactive, PIN-gated web microsites — not PDF decks. The roadmap included a designed proposal for a recurring observation-agent phase as the logical next step.
The solution.
A consult engagement that ended with the practice knowing exactly what to do next, in what order, and what to skip entirely.
PIN-gated discovery web app
38 questions, 10 sections, autosaving, structured responses straight into our CRM — discovery as a product, not a PDF.
Dual-EHR workflow map
A documented picture of where work duplicates between the two systems, including the ~5 hrs/week of manual re-keying in one admin role.
Ranked AI opportunity map
Every candidate use case scored on impact, complexity, and compliance risk, with infeasible automations flagged rather than pitched.
Marketing visibility audit
SEO scored at 3.2/10, AI-assistant visibility at 2.4/10, with 16 local competitors mapped for context.
EHR consolidation research
Two independent expert research passes to inform a platform consolidation decision.
Interactive microsite deliverables
Everything delivered as PIN-gated web microsites the team can revisit — not a slide deck that dies in an inbox.
Results & ROI.
This was a consult, so what we can honestly claim is an operating picture and a prioritized roadmap — not measured savings. The recurring observation-agent phase we designed has been proposed, not installed, and we won't report results from work that hasn't shipped.