Philly hospital networks have 8-figure EHR budgets but zero budget for the glue code between EHR, billing, and patient comms.
Pennsylvania · AI Automation Services
Custom n8n pipelines for Philly healthcare systems, Pittsburgh manufacturers, and PA education tech.
Philly hospital networks have 8-figure EHR budgets but zero budget for the glue code between EHR, billing, and patient comms.
Pittsburgh mid-cap manufacturers run quality control via clipboard and Excel — defect traceability takes days, not minutes.
PA EdTech and university ops teams manually reconcile registration, payment, and LMS data across three legacy systems.
Client snapshot
A Pittsburgh specialty manufacturer was losing ISO audits because defect data lived in paper QC logs, Excel sheets, and a shift supervisor's head. I built an n8n flow that ingests mobile-photo QC cards via GPT-4o vision, classifies defects, logs to a central DB, and generates audit-ready reports. Audit prep went from 3 weeks to 2 days.
Who builds it: Waseem Nasir — AI automation engineer, digital nomad from Pakistan, based in Bali/ASEAN, serving global clients.
Not directly on HIPAA-protected workflows without a BAA and a security review. But I build on self-hosted n8n with HL7/FHIR connectors for non-PHI layers — patient reminders, scheduling, non-clinical comms. Clinical data stays in Epic.
Scope drives cost. A single QC-automation workflow runs in the 40-60 hour range; a full plant-wide system is 4-8 weeks. Fixed-fee SOWs with weekly demos keep the budget predictable — no scope creep, no surprise invoices.
Clinical-trial document intake. GPT-4o vision and Claude Opus both parse consent forms, eCRF exports, and lab PDFs into structured data 10x faster than manual data entry — and with a validation layer, error rates drop below human baseline.
NDAs — always, before the first discovery call if needed. BAAs — only where my tooling actually touches PHI, which I try to avoid by design. For most workflows I keep PHI on client infrastructure and automate around it.
30-minute discovery call. No pitch deck. I map your current workflow, identify the highest-leverage automation targets, and tell you whether it makes sense to work together.