First written in March 2026 and updated in October 2026 to reflect the program as it stands.
I joined Right at Home in 2025 as its first full-time AI hire, and there was no AI strategy yet. I wrote the company's first AI strategy and three-year roadmap. This is the order we did it in, and the decisions that got products shipped instead of stuck.
Phase 1: The Strategy Before the Technology
Before building anything, we wrote the AI strategy document, and it went through several versions as we learned what the organization would accept and what its users needed.
Right at Home is a 480+ location in-home senior-care franchise network, and the corporate office supports all of them. The first use case served the corporate office.
Phase 2: Governance as a Permission Structure
I stood up the Enterprise AI Governance Council, and we built out the compliance policy framework. Governance can feel like overhead when there is pressure to ship, but it is what let every later deployment move faster.
The practical outputs of the governance work were a vetting process for AI tools and vendors and safe-adoption guidelines for the organization. Together, these answered the question that would otherwise block every future deployment: is this tool approved?
Governance gave people a path to yes. Employees who want to try an AI tool now have a defined process with a defined answer, and the same documentation answers a business unit's vendor pilot or a legal question about data handling.
The council drew its members from across the business units, so the policy carried weight outside the technology team.
Phase 3: Two Corporate Products, Then a Franchise Track
The first two products served corporate staff, two distinct problems with two distinct answers.
Self-service data analytics. The program's first use case was an analytics agent on Snowflake Cortex that translates plain-English questions into queries against our data warehouse. The target user was a corporate employee who knew the business deeply but wasn't a data analyst. An operational question used to mean filing a data request and waiting on a report. The analytics agent lets them ask in plain English and get a number back.
Before building, I evaluated the available platforms that turn plain-English questions into database queries. The team then built the capability in-house for a small fraction of what outside vendors quoted each year: the analytics developer, a technology lead and I shipped it together, piloting inside the corporate team before a wider rollout so we could tune answer quality before the stakes were higher.
Knowledge retrieval via Microsoft 365 Copilot. The second product served corporate staff who needed reliable answers to operational questions: policies, procedures and compliance requirements spread across many documents and internal resources. The information existed, but finding it took too long and too often turned up the wrong version.
Built with the company's technology team, it's a corporate knowledge agent on Microsoft 365 Copilot and Azure AI Search that answers corporate staff's questions from the company's policy and operational documents.
Deploying through Microsoft 365 Copilot met users where they already worked instead of asking them to adopt a new tool, which removed the usual first hurdle to adoption.
Then two tracks. The success of that first use case widened the AI roadmap from the corporate office alone to two tracks, giving the franchise network and the corporate office equal weight in deciding what gets built next. The franchise track is where the third product came from: a voice AI sales-coaching platform for the franchise network's sales representatives.
Phase 4: Enablement
Shipping a product does not change how people work, so the program counts adoption, not launches, as the result.
The enablement program seats an AI advocate in each corporate function and runs a standing schedule of enablement sessions; the governance post covers how both work.
The franchise network gets its own enablement: franchise education sessions, readiness for the corporate coaches who introduce new products to franchise owners, and the guides, FAQs, office challenges and onboarding that move offices from awareness to everyday use. I led that franchise-facing work and co-led the enablement strategy across the corporate office and the franchise network, including the company's AI Strategy Playbook.
The advocates mattered most: staff who know the friction in their own work find the best use cases, and the advocates feed those straight into the roadmap.
The result: teams across the corporate office now build and run agents of their own.
What I Took From It: AI Strategy Is Product Management
Looking back at the full arc (strategy, governance, two products, enablement), the frame that best describes the work is product management.
I came to see the AI strategy as a product, with the organization as its user. That meant user research before building, shipping in versions as we learned, defining success before launch, and treating governance as the feature that lets people rely on the tools.
It also means accepting that the roadmap will change. The two products we shipped first reflected where the organization was when we started, and the roadmap keeps changing as the organization learns and the technology moves; each solved problem exposes the next one.
A year and a half in, looking at every request through both tracks, corporate and franchise, is second nature for the team.
When this was first written, the Right at Home portfolio was these two products. A third, Connect, a voice AI sales-coaching platform where franchise sales representatives rehearse calls against AI buyer personas, was piloted with franchise offices through the summer of 2026 and is rolling out across the franchise network, office by office as each signs its agreement.