New State of AI 2026: Mid-Year Reality Check is live. Read the report
Skip to content
Case Study K9 Resorts

From Disconnected Systems to Conversational Intelligence Across 52+ Franchise Locations

K9 Resorts now has a centralized, enterprise-grade data platform that unifies operational and financial data across 52+ franchise locations for the first time.

Client K9 Resorts
Sector Dog boarding and daycare franchise
Network 52+ franchise locations
K9 Resorts luxury pet hotel, the franchise network unified on a single data platform by HatchWorks AI
Hatchworks Team Data Engineers, AI/ML Engineers, Solution Architect, Product Owner, Project Manager
Overview

A theory worth testing: scale the network, hold corporate costs flat

K9 Resorts is a fast-growing, private-equity-backed franchise organization. It came to HatchWorks AI with a theory: that AI and data solutions could help it scale the business while holding corporate costs flat and improving franchise performance. To test that theory, K9 Resorts set out to unlock data-driven insights across its 52+ franchise locations through an enterprise-grade data platform, and HatchWorks AI delivered two integrated products.

A K9 Resorts luxury pet hotel location, one of the 52+ franchises unified on the data platform

K9 Fetch

K9 Fetch is a real-time analytics dashboard covering financial performance, occupancy, forward booking curves, and cross-franchise benchmarking.

K9 Chat

K9 Chat is an AI-powered assistant that lets operators and corporate users query franchise data in plain English.

Together, they move K9 Resorts from manual reporting to intelligent, self-service decision support at scale.

The Challenge

Two systems, no integration layer, no unified view

K9 Resorts operates 52+ dog boarding and daycare franchise locations across the United States with no unified view of business performance. Operational data lived in Gingr. Financial data lived in ProfitKeeper. The two systems had no integration layer between them.

What K9 Resorts was up against:

  1. 01

    No real-time visibility.

    Corporate leadership had no live view of KPIs across the network.

  2. 02

    No benchmarking.

    There was no way to compare franchise performance against peers.

  3. 03

    No AI foundation.

    The business had no technical base for AI-driven decision making at scale.

  4. 04

    Manual, time-intensive support.

    Regional Directors relied on slow manual processes to support franchisees.

  5. 05

    An access gap.

    Non-technical operators, franchise owners, regional directors, and corporate staff needed insights without navigating complex dashboards or writing queries.

The challenge was not just unifying the data. It was making that data accessible and conversational for the people who need it most.

The Process

One platform, then two products on top of it

HatchWorks AI designed and built an enterprise-grade data platform on Databricks, then layered two products on top of it.

How the data flows

Two source systems in. One validated Gold layer in the middle. Two products out, both reading the same numbers.

Step 01 Sources
Gingr Operational data lived in Gingr.
ProfitKeeper Financial data lived in ProfitKeeper.
Step 02 Ingestion
Nightly pipelines Unify operational data from Gingr and financial data from ProfitKeeper across all franchise locations.
Step 03 Lakehouse on Databricks
Bronze Raw ingested data.
Silver A reservation-explosion transformation layer.
Gold Clean, validated daily-grain KPI aggregates.

Built as a Bronze/Silver/Gold lakehouse architecture on Databricks. Revenue figures are validated to penny-level accuracy against Gingr source data.

Step 04 Serving layer
Google Cloud SQL Analytics layer.
Custom REST API Serves validated figures to K9 Fetch.
Semantic API index An indexed catalog of all validated backend endpoints, so the assistant uses the same business logic that powers K9 Fetch.
Step 05 Products
K9 Fetch Financial P&L, occupancy, forward booking, and cross-franchise benchmarking, with role-based access control.
K9 Chat Plain-English questions answered through RAG and Databricks Genie against the validated Gold layer.
  • Built a Bronze/Silver/Gold lakehouse architecture on Databricks.

  • Created nightly ingestion pipelines that unify operational data from Gingr and financial data from ProfitKeeper across all franchise locations.

  • Built a reservation-explosion transformation layer that produces clean, validated daily-grain KPI aggregates.

61% answerable
KPI feasibility assessment

The team also ran a structured KPI feasibility assessment across 70+ question types and confirmed that 61% are answerable directly from the current platform.

The Outcome

From static dashboards to answers on demand

K9 Resorts now has a centralized, enterprise-grade data platform that unifies operational and financial data across 52+ franchise locations for the first time.

K9 Fetch delivers real-time dashboards covering financial performance, occupancy, forward booking, and benchmarking, giving corporate leadership and regional directors a live view of the entire network.

Revenue figures are validated to penny-level accuracy against Gingr source data across all service categories.

K9 Chat brings those same validated figures into a natural language interface. Operators and corporate users ask questions in plain English and receive answers grounded in the identical data that powers K9 Fetch.

The result moves K9 Resorts from static dashboards to conversational, on-demand business intelligence, with no technical expertise required from end users.

Corporate leadership had no live view of KPIs across the network.

There was no way to compare franchise performance against peers.

The business had no technical base for AI-driven decision making at scale.

Regional Directors relied on slow manual processes to support franchisees.

Non-technical operators, franchise owners, regional directors, and corporate staff needed insights without navigating complex dashboards or writing queries.

Key Stats

The network, in numbers

52+

franchise locations unified on a single data platform

2

primary data sources integrated (Gingr and ProfitKeeper)

70+

natural language KPI question types assessed

61%

of KPI questions answerable directly from the current platform

  • Revenue validated to penny-level accuracy across all service categories

  • Role-based access control across corporate, regional, and franchise tiers

  • AI assistant answers grounded in the same validated data as the K9 Fetch dashboard

One platform, every location

Each mark is a franchise location. The final mark stands for continued network growth.

Solution at a Glance

Five pieces, one system

01

Enterprise data lakehouse on Databricks (Bronze/Silver/Gold), fed by nightly Gingr and ProfitKeeper ingestion pipelines

02

K9 Fetch real-time analytics dashboard: financial P&L, occupancy, forward booking, and cross-franchise benchmarking

03

K9 Chat AI assistant using RAG and natural language to SQL via Databricks Genie

04

Semantic API index for dashboard-accurate AI responses

05

Role-based access control across corporate, regional, and franchise tiers

Technologies Used

The stack

Databricks (Bronze/Silver/Gold lakehouse) Gingr (operational source) ProfitKeeper (financial source) Google Cloud SQL (analytics layer) custom REST API React (K9 Fetch dashboard) Claude Sonnet by Anthropic (LLM) LangChain (AI orchestration) retrieval-augmented generation (K9 Chat) Databricks Genie (natural language to SQL) vector embeddings semantic API index
Lessons Learned and Next Steps

Design the data layer for dashboards and for AI, from the outset

This engagement showed the value of a unified data lakehouse for multi-location franchise operations. It also showed the importance of designing data layers for both dashboard consumption and AI querying from the outset. The Gold layer, first optimized for dashboard aggregations, needed a complementary API index to enable deterministic, dashboard-accurate AI responses.

A key architectural insight: when building AI assistants over operational data, grounding responses in validated, already-trusted business logic produces more reliable results than generating SQL from scratch.

Next steps include:

  1. Next 01

    Expanding K9 Fetch to additional KPI categories.

  2. Next 02

    Scaling the semantic API index as new dashboard endpoints are added.

  3. Next 03

    Exploring Databricks Unity Catalog metric views as a more flexible query layer for K9 Chat.

  4. Next 04

    Deepening question coverage through structured engagement with K9 operations SMEs.

  5. Next 05

    Scaling the platform to support the full K9 Resorts franchise network as it grows.

Client Quote

In their words

“For the first time, we have a single, trusted view of performance across every K9 Resorts location. That changes how we make decisions as a brand.”
Scott Troeller, CEO, K9 Resorts
“What sold me was the accuracy. The numbers in K9 Chat match the dashboard to the penny because they come from the same validated logic.”
Kevin Tennant, Data & Analytics Lead, K9 Resorts
Context

About the work

About K9 Resorts

K9 Resorts Luxury Pet Hotel is an award-winning dog boarding and daycare franchise founded in 2005 by brothers Steven and Jason Parker in Fanwood, New Jersey. The brand began franchising in 2011 and now operates a growing network of luxury pet hotels across the United States. Each location pairs cage-free luxury boarding with hospital-grade ventilation, antimicrobial flooring, and professionally trained staff. K9 Resorts is one of the most awarded brands in the pet care industry, recognized multiple times by the International Boarding and Pet Services Association.

About HatchWorks AI

HatchWorks AI turns AI into ROI by automating key business processes, transforming data, deploying intelligent agents, and shipping AI-powered products that deliver measurable results.

Ready to make your data conversational?

See how HatchWorks AI unifies fragmented systems into a platform your teams can simply ask questions of.