AI Agent Operational Lift for Fetch Inc (now Fetch Pet Insurance) in New York, New York

Why AI matters at this scale

Fetch Inc. (formerly Petplan) operates in a sweet spot for AI transformation. With 201-500 employees and an estimated $45M in annual revenue, the company is large enough to have meaningful data assets and process pain points, yet small enough to adopt AI without the bureaucratic inertia of a legacy carrier. The pet insurance market is growing at over 15% CAGR, but margins are squeezed by manual claims handling and customer acquisition costs. AI can be the lever that turns this mid-market player into a category leader.

What Fetch does

Fetch provides accident, illness, and wellness insurance for dogs and cats through a direct-to-consumer digital platform. Customers submit claims by uploading veterinary invoices, which are manually reviewed by adjusters. The company competes on coverage breadth and customer experience, but the back-office remains heavily human-dependent. This creates a clear AI opportunity: automate the ingestion, interpretation, and adjudication of claims to reduce cost and improve speed.

Three concrete AI opportunities with ROI framing

1. Intelligent Claims Automation. The highest-impact use case is applying natural language processing (NLP) and optical character recognition (OCR) to veterinary invoices. Today, adjusters manually transcribe line items, check policy terms, and calculate reimbursements. An AI pipeline can extract procedures, medications, and costs with high accuracy, auto-approve claims that fall within clear parameters, and flag only exceptions for human review. For a company processing hundreds of thousands of claims annually, reducing manual touch by 60-70% could save $2-3M per year in operational costs while cutting reimbursement time from days to hours—a direct driver of Net Promoter Score and retention.

2. Predictive Underwriting. Pet insurance pricing has traditionally relied on broad actuarial tables. Machine learning models trained on breed-specific morbidity data, geographic risk factors, and historical claims can generate personalized premiums that better reflect true risk. This reduces adverse selection and improves loss ratios. Even a 2-3 point improvement in loss ratio on a $45M book translates to nearly $1M in additional underwriting profit annually.

3. AI-Driven Customer Engagement. A conversational AI layer—deployed on web and mobile—can handle policy inquiries, coverage lookups, and claim status checks 24/7. This deflects routine support tickets, freeing agents for complex cases. When combined with a recommendation engine that suggests wellness add-ons at renewal, the AI stack can lift average revenue per user by 5-10% while reducing service costs.

Deployment risks specific to this size band

Mid-market insurers face unique AI risks. First, talent scarcity: competing with tech giants for ML engineers is difficult, so Fetch should consider managed AI services or low-code platforms. Second, data quality: veterinary invoices are unstructured and inconsistent; models require robust training data and human-in-the-loop validation to avoid errors that could trigger regulatory scrutiny. Third, integration: core systems like policy administration and claims management may not have modern APIs, necessitating middleware investment. A phased rollout—starting with claims automation as a standalone microservice—limits blast radius and builds internal confidence before expanding to underwriting or customer-facing AI.