PawCalm.
By Yash Shanmughanandam
AI behavioral concern assessment for dog owners — anxiety-reduction first, decision support (Monitor / Try This / Call Vet) grounded in personalized dog profiles.
The problem
Dog owners experience an emotional escalation pattern that existing tools make worse. Owner notices behavioral change (dog not eating, unusual barking, lethargy, aggression) → anxiety spiral begins ("Is this normal? Is something wrong? Am I a bad pet parent?") → Google search worsens the anxiety (worst-case results dominate) → decision paralysis ("Should I wait? Should I call the vet? Is this an emergency?") → expensive vet telehealth call for what's often a benign behavioral change.
The pain stack is brutal and recurring: anxiety spiral (severity 10/10, high frequency), decision paralysis (9/10), Google rabbit holes, no trusted source for perspective, and the additional guilt of feeling like a "bad pet parent" for being uncertain.
The solution
PawCalm sits explicitly between Google panic and vet calls. The owner logs a concern through a 5-step guided wizard (under 2 minutes); the AI analyses the concern against the dog's personalized profile (breed, age, health history, behavioral baseline); returns structured guidance with a clear recommendation: Monitor (likely benign, watch for X), Try This (specific behavioral or environmental adjustment), or Call Vet (warrants professional input, here's what to tell them).
The product's distinctive design choice is anxiety reduction as the primary goal, not information density. The AI uses the dog's name, validates the parent's concern, and frames the response with reassurance even when the recommendation is to escalate.
How it works
Flow: trigger (behavioral change observed) → open PawCalm → tap "Log a Concern" (prominent CTA on home screen) → 5-step guided input (what behavior, when did it start, duration, dog context, optional photos) → AI assessment against dog profile → results display with clear recommendation (Monitor / Try This / Call Vet) → optional save to history.
System prompt: positions the AI as a supportive assistant, NOT a veterinarian — cannot diagnose medical conditions or recommend medications. Tone: warm, empathetic, reassuring; uses the dog's name; validates the owner's concern before guiding.
Primary model: Anthropic Claude 3.5 Sonnet, chosen for strong reasoning for pattern interpretation, excellent structured JSON output, built-in safety consciousness for medical-adjacent use cases, and consistent persona maintenance. 50+ test cases cover typical scenarios (won't eat breakfast, scratching, lethargy, barking at nothing).
Who it's for
Primary persona — the "Worried Pet Parent": urban/suburban dog owners aged 28–45, household income $75K+, already spending on premium pet products, consider their dog a family member, and experience real guilt/anxiety about their pet's wellbeing. They've Googled symptoms more than once and felt worse for it.
B2C primary (direct-to-consumer mobile app). B2B2C future through partnerships with pet insurance providers (decision support reduces claim escalations), veterinary clinics (triage tool that reduces "is this an emergency?" calls), and pet retailers (engagement driver).
Why it matters
97% of pet owners consider pets family members. The US pet industry is ~$150B at 6–8% YoY growth. Pet Tech is at 15–20% CAGR through 2028, with AI-powered pet solutions growing 25%+ annually and pet health & wellness at ~$35B with 8–10% YoY growth. The pet humanization trend is structural and accelerating, and the post-pandemic remote-work shift has owners spending more time with their pets — and noticing more behavioral changes.
PawCalm's structural bet is on anxiety-reduction as a category position, not symptom-checking. Symptom checkers are a commodity; trustworthy decision support that learns each dog over time is much harder to copy and much more valuable to the user. Once a Worried Pet Parent has logged a few concerns and the AI's recommendations have proved out, the product becomes the default first stop for every future behavioral change.
At a glance
- Project
- PawCalm
- Built by
- Yash Shanmughanandam
- One-liner
- AI behavioral concern assessment for dog owners — anxiety-reduction first, decision support (Monitor / Try This / Call Vet) grounded in personalized dog profiles.