AI Agent Operational Lift for Ottobock Care (wright & Filippis) in Rochester Hills, Michigan

Why AI matters at this scale

Wright & Filippis, operating as Ottobock Care, is a 200-500 employee regional powerhouse in home health and durable medical equipment (DME). Founded in 1944, the company provides custom prosthetics, orthotics, and respiratory therapy across Michigan. At this size, the organization is large enough to generate meaningful proprietary data from thousands of patient encounters but often lacks the massive IT budgets of national hospital chains. AI is the great equalizer here—it can automate the high-touch, documentation-heavy workflows that eat into clinician time and compress margins in a reimbursement-constrained industry.

The core business: High-touch care, high-volume data

The company's model blends clinical services with product fulfillment. Clinicians fit patients for custom orthotics or CPAP masks, document the encounter, and trigger a complex supply chain of fabrication and resupply. This creates a rich longitudinal dataset: 3D scans, fitting notes, refill histories, and insurance interactions. For a mid-market firm, this data is an underutilized asset. AI can turn it into a predictive engine that anticipates patient needs before they call, reducing the churn that silently erodes revenue.

Three concrete AI opportunities with ROI

1. Ambient clinical intelligence for immediate margin relief. Prosthetists and respiratory therapists spend up to 40% of their day on documentation. Deploying an ambient AI scribe that listens to patient visits and drafts structured notes in the EHR can reclaim 5-8 hours per clinician per week. For a staff of 50 clinicians, that's the equivalent of adding 5-6 full-time practitioners without hiring—a direct labor cost savings of $300k-$400k annually.

2. Predictive resupply and adherence modeling. CPAP and oxygen patients require regular mask replacements, tubing, and filters. Missed resupply windows represent a significant revenue leakage. An AI model trained on historical refill patterns, patient demographics, and seasonal trends can flag at-risk patients for proactive outreach. A 15% improvement in resupply adherence could add $1M+ in annual recurring revenue with near-zero marginal cost.

3. AI-assisted custom orthotic design. The transition from plaster casting to 3D scanning opens the door for generative design algorithms. AI can analyze a library of successful socket or orthotic designs and suggest an initial model based on a new patient's scan and diagnosis. This cuts design time by 30-40%, allowing senior clinicians to focus on fine-tuning and patient interaction rather than CAD groundwork.

Deployment risks for the 200-500 employee band

The primary risk is integration complexity. Mid-market DME providers often run on legacy or niche platforms like Brightree, where APIs may be limited. A phased approach is critical—start with a standalone AI scribe that doesn't require deep EHR integration, then move to supply chain models once data pipelines are proven. Change management is the second hurdle; clinicians will resist tools perceived as surveillance. Success requires positioning AI as a "co-pilot" that eliminates paperwork, not a replacement for clinical judgment. Finally, data governance must be airtight. Patient data used for model training must be de-identified and processed in a HIPAA-compliant environment, which is achievable with modern cloud services but demands upfront investment in security architecture.