PawHealthAI: product design and engineering
PawHealthAI is a pet-care app that helps pet owners organise care and explore next steps through Evie, an AI assistant offering conversational guidance, structured symptom triage, weekly wellbeing check-ins, reminders and health records. It gives non-diagnostic guidance and does not replace a veterinarian.
- Co-Founder & CTO
- Akash Renuka Ashok
- Primary market
- Australia
- Users
- 189 registered users (October 2026; not active usage)
- Launched
- 2025
- Redesign relaunched
- April 2026
What PawHealthAI does
- Evie, an AI assistant for conversational pet-health guidance
- Structured symptom triage with urgency routing
- Weekly wellbeing check-ins and longitudinal pet context
- Reminders and pet health records
Is PawHealthAI a diagnosis tool?
No. Evie is deliberately non-diagnostic: concern classification, urgency routing, prompt-injection checks, post-generation sanitisation and hard constraints prevent it from diagnosing or prescribing, and it directs owners to a vet when a concern is urgent.
How it was built
- Designed and built a pet-care platform with 189 registered users as of October 2026; Australia is the primary market, and registrations do not measure active usage
- Architected Evie across conversational support, structured symptom triage, weekly wellness check-ins and the Pawport longitudinal pet record
- Designed non-diagnostic safety controls: concern classification, deterministic emergency escalation, medication guardrails, prompt-injection checks, post-generation sanitisation and hard constraints against diagnosis or prescription
- Cut pet-history retrieval time from ~700ms to ~100ms, measured in development, by running independent database requests concurrently instead of sequentially
- Secured distribution through Beny, an employee-benefits platform with 32,000 eligible members, with member distribution confirmed through CopCare (13,000+ members) and a 70-handler cohort onboarding with Working Paws Australia
- Built a dedicated KPMG employee pet-care benefit offer; completed the SMEC AI Accelerator and MEC Discover
- Shipped Notification System V2, combining device-local scheduling with server-sent contextual push
Stack: React Native 0.81 (Expo SDK 54), React, Vite & Tailwind CSS, TypeScript & Node.js, Firebase Cloud Functions (Gen 2), Supabase PostgreSQL (RLS), Google Gen AI SDK (Gemini 2.5 Flash), RevenueCat, Sentry.
Explore the product
Case studies
- PawHealthAI: product design and engineering
Home-screen redesign, onboarding evolution and reusable React Native components for a pet-care product. - Redesigning PawHealthAI after launch
What users told me about version one, and the simpler app I rebuilt in response for the April 2026 redesign. - PawHealthAI onboarding evolution
Six steps, then two plus four, then completed-page saving. The decisions, trade-offs and remaining validation.
Last updated