About the Company
Our client is a US-based healthcare outreach automation platform that helps care teams communicate with patients at scale, using AI-driven SMS and Voice workflows. Patient outreach - reminders, follow-ups, scheduling, redeterminations - is the silent operational tax of running a healthcare practice. The product's premise is that most of this work can be automated safely, freeing care teams to focus on the patients and decisions that need human judgment.
Tequity has been the client's design and engineering partner since the engagement began, owning end-to-end product development as the platform scales across customer tenants.
The Challenge
A healthcare communications product has to clear a high operational and clinical bar:
- Multi-modal outreach: SMS and Voice flows have to feel coherent, work across 90+ patient languages, and handle responses correctly the first time.
- AI auto-response with safety: Automated replies have to interpret patient intent (time preferences, follow-up willingness) without overstepping clinical context.
- Operational tooling for care teams: Care coordinators need a workflow surface for follow-ups, voicemails, templates, and patient activity that reflects real practice operations.
- Multi-tenant complexity: Different practices have different working hours, workflows, and external system integrations. The platform has to support this without forking code.
- Observability and analytics: AI workflows in healthcare require traceability for every decision and message.
Solution Implemented by Tequity
Tequity owns design and engineering across the platform.
AI Outreach Engine:
- AI auto-response improvements for SMS conversations, including time-preference detection and automated replies for common patient responses.
- Operational knowledge configuration so AI replies incorporate clinic working hours and contact information.
- Voice AI follow-up task reliability and improved task generation prompts for more accurate care team actions.
- VoiceAI enhancements: transcript access, rescheduling flows, and improved call lifecycle handling.
Care Team Tooling:
- Follow-Ups page redesigned with customizable columns, task assignment, notes, quick counts, and voicemail tracking.
- Template management for admins, with inline editing, language-based grouping, and a simplified structure.
- Patient updates log with centralized notes, activity tracking, and appointment updates against a unified data model.
Patient Workflows:
- Patient follow-up management with new fields (Reporting Type, Renewal Date), enhanced filtering, sorting, and search.
- Auto-enrollment in automated sequences based on outreach reason and language, with fallback handling.
- Redetermination workflow with renewal status, scheduling, and full sync with follow-ups.
- Patient language standardization across 90+ languages with spelling-variation handling.
Integrations and Infrastructure:
- Optexity integration for PCP-change automation against external insurance portals, with one-click action and progress tracking.
- Langfuse trace integration for AI features - exporting traces with evaluation labels and notes, scoped by tenant, into Google Sheets.
- Translation pipeline so outbound messages and VoiceAI transcripts now support automatic English translations.
- Aurora Serverless DB, ETL pipelines, audit logs, and scheduled data processing for analytics infrastructure.
- Multi-patient phone number support, real-time task updates, and cross-tenant clarity enhancements for system scalability.
Platform Hygiene:
- Removed unused modules (chatbot, metrics dashboard, member roster) to simplify the system.
- Automated checks added to improve reliability and maintainability.
Results
The platform is actively scaling across customer tenants:
- 100,000+ AI-generated SMS sent through the platform
- 10,000+ Voice minutes handled by VoiceAI flows
- 90+ languages supported in patient communication
- Multi-tenant with per-tenant configuration, traceability, and analytics
