Frequently asked questions

Three services are built as a first step: the half-day AI Leadership Workshop ($3,500), the two-week AI Readiness Sprint ($9,500) and, if your hospital is in CMS's TEAM model, the free TEAM position report. A small practice can start with the owner's version of the Weekly Leadership Brief ($6,500 to build, then $250 a month) or the $1,500 website and reviews check.

If you already know the problem, describe it on a call or in the written form, and we'll tell you which service fits, or that none does. For a recommendation without talking to anyone, take the six-step quiz: it names one service and two alternates, with prices and the reasons for each, and asks for no contact details.

Yes, if you're in healthcare. The services are built for health systems; community hospitals, including rural and critical access hospitals; health centers, free and charitable clinics and local health departments; physician, dental and behavioral health practices; post-acute and senior care, including skilled nursing, home health, hospice, senior living and PACE; and health-tech companies, revenue cycle and staffing firms, associations and medical societies. The same methods work in any industry with data and a workflow problem.

No. Fill in the written form, which takes about five minutes, and within two business days you get a written recommendation with a fixed price, or the few questions Tyler needs answered first. Please leave patient information out of it. Or book a 20-minute call, and if one of our services fits, a two-page proposal with a fixed price follows within 48 hours.

Either way, the proposal lists what gets delivered, the schedule, what we need from your team, how data is handled and the price, with written acceptance criteria and an expiry date.

Your proposal sets the start and finish dates in writing. Each service's page says how long the work takes, from a 90-minute briefing for leaders to 6 to 12 weeks for a custom app.

Yes. Many services have a smaller first step, and its fee counts toward the larger service if you book that within 90 days. A 90-minute executive briefing for leaders ($1,800) counts toward the AI Leadership Workshop. The $3,500 workshop fee comes off an AI Readiness Sprint, the AI Governance Starter or a first month of the Fractional AI and Data Lead. The $9,500 Sprint is credited toward a Custom AI Tool build.

Each AI tool has a smaller one-time step, such as a check, audit or pilot, that counts toward its build, and the $14,500 Self-Service Analytics pilot counts toward the full build. Two first steps are priced on their own: the Capacity and Staffing Forecast's trial run and Spreadsheet to App.

Several services have a set point where the work can stop early. On the AI Readiness Sprint and the Throughput Diagnostic, if we find in the first three days that the work won't pay off, we stop and return your deposit. Predictive Models stop after week one if your data can't support a model, with nothing due beyond the signing payment. The Capacity and Staffing Forecast starts with a trial on your own past months, and the work goes no further if the forecast doesn't pass it.

On every service, nothing counts as finished until it works on real examples from your own organization. And if AI or data work can't fix your problem, Tyler says so on the first call.

Yes. Hospitals in CMS's TEAM model, which sets Medicare spending targets for five surgeries, can get a free TEAM position report that compares their costs with those targets. It's built from public CMS files, with nothing to sign and no data to send. Nursing homes can ask for a free one-page snapshot of their citations and fines against the rest of their state.

The quiz and the 20-minute call cost nothing, and neither does anything in Insights, where the charts and facts are free to share and embed with credit to Kent Applied.

No. Every service has a fixed price or a flat monthly fee, written into your proposal before work starts. Where the menu shows a range, the proposal sets the exact figure, and each service's page says what decides it, such as the number of source systems or vendors. Anything outside the written scope is priced in writing before it starts.

Some costs sit outside our price. Where the work runs in your own accounts, the cloud, hosting and AI usage bills come from those providers straight to you, and the proposal estimates them before you sign. For the workshop and training, travel beyond San Diego County is billed at cost.

Under $20,000, you pay half at signing and half at delivery. From $20,000, it's 40% at signing, 30% at a midpoint milestone and 30% at acceptance, when the work meets the written criteria in your proposal.

Purchase orders are fine, and with one there's no deposit: we invoice half at kickoff and half at acceptance, or from $20,000, 25% at kickoff, 35% at the midpoint and 40% at acceptance. The Fractional AI and Data Lead is billed monthly in advance, by invoice or purchase order, with a three-month first term and then 30 days' notice.

Paying from a federal grant? Some AI tool versions stay under the $15,000 federal micro-purchase limit for year one, so they can be bought without competing quotes if you document that the price is reasonable and your own purchasing rules allow it. Each tool's page says which versions qualify.

Each tool has a one-time build price, then a flat monthly fee, month to month with 30 days' notice. The fee covers checking every run, fixing the tool when a source or system changes, and replies by the next business day. Some versions have no fee.

A tool that reads your files runs in your own accounts, which we set up with you during the build. You pay its hosting and AI usage directly, and the proposal estimates that cost before you sign. If you stop our fee, the tool keeps running and stays yours. A tool that reads only public information runs on our servers, with those costs in the fee; if you stop, you get every template, draft and finding as files.

With three or more of our tools, the monthly fees are capped at $1,000 in total. Health system, network and multi-site versions are priced on their own.

For data, analytics and app builds, nothing goes to us unless you choose support. Every build ends with documentation, a recorded training session and 30 days of fixes after acceptance. Optional support after that costs $300 to $1,500 a month, month to month, depending on what was built. Without it, your team has the code, the documentation and the training to run what we built.

Cloud, warehouse and hosting bills go to your own accounts, and the proposal estimates them before you sign. The AI tools have their own monthly fee.

For most of the work, no. Seven of the eight AI tools are built to run without it, and a Custom AI Tool that would read patient records starts only once a BAA and our insurance are in place. Whether you still want a BAA for an AI tool is your compliance officer's call, so raise it before signing. Of 44 common healthcare problems we mapped, an AI tool can help with 33 without reading any patient data.

A few services need patient-level data, such as the admission and discharge times the Throughput Diagnostic uses. Work on patient-level data starts once a BAA or Data Use Agreement and our insurance are in place. That includes a limited data set, meaning records with names and record numbers removed that can still show dates and ZIP codes. Until then, the work uses counts, summaries or de-identified data. To hear when that changes, say so when you ask for a written proposal.

Wherever possible, your data stays in your own systems. Tools that read your files run in your own Microsoft 365, Google Workspace or cloud account, with the AI model on an account in your name, and nothing is copied to our systems. Our support login sees only the folders the tool reads. Only tools that read public information run on our servers.

Your statement of work names every AI tool and what it can see, and working copies made during an engagement are deleted at the end.

No. We use only business accounts set so the AI company never trains its models on your data, never the free or personal accounts anyone can sign up for.

No AI tool acts on its own. Each one does one job on a schedule, and a person you name approves what it produces before anything is sent, posted or changed. Any observation the AI adds is labeled as AI, shows what it's based on and appears separately from facts pulled from your systems. In the Weekly Leadership Brief, ordinary code calculates every figure and links it to its source; the AI writes only the sentences around the numbers and can't change them.

Before handoff, each tool has to meet the written "What done means" criteria on its page, such as a set number of runs reviewed and approved, with every figure traced to its source.

Yes. Send your questionnaire and Tyler will answer it. Work that runs inside your own systems and accounts stays under your own controls, and each statement of work says what data is used, where it stays and which AI tools touch it.

Tyler Kent does it himself, and there is no junior team to hand your project to. He has 13+ years in data science and analytics, most of it in healthcare, and has designed and shipped 10+ AI products to production. He uses AI development tools under his own review, and he is responsible for every deliverable.

An AI assistant drafts replies to requests and first drafts of proposals, and Tyler reviews and approves each one before it goes out. The one automatic message is a short note from desk@kentapplied.com confirming that your request arrived. The privacy page explains how this works, and Tyler's background is on the About page.

Yes. The Portfolio shows four AI systems Tyler designed and built for a public data platform on healthcare and aging, with identifying details withheld. His background and nine peer-reviewed journal papers are on the About page.

Anyone can open a sample TEAM position report, a sample Weekly Leadership Brief and the research and charts in Insights.

About three hours of interviews in week one, spread across two or three people. You name one point of contact who can make decisions, and one person who owns the measure the work is meant to improve. After that, we need system access or data extracts by the dates in the statement of work, and answers within two business days; we send our questions in batches.

Usually. We build on the systems you already run, such as Epic, Oracle Health or MEDITECH for admission and discharge times, Snowflake, BigQuery or Databricks for warehouses, and Power BI, Tableau or Looker for dashboards. AI tools that read your files run inside your own Microsoft 365, Google Workspace or cloud account, and they send their results by email, Teams or Slack, or into software your team already uses. Each service's page names the systems it works with.

Kent Applied is based in San Diego, and the work can be done remotely for an organization anywhere in the US. On-site sessions are by arrangement, when a kickoff, workshop or readout works better in the room. For the workshop and training there's no travel charge in San Diego County, and travel beyond it is billed at cost.

You do. Once the final payment is made, the deliverables, the code and every output are yours, and your data is always yours. General-purpose parts we reuse across projects are licensed to you permanently for your own use.

There are two exceptions. Kent Applied keeps the training curriculum, and your team gets a permanent license to use what was delivered. And for an AI tool that runs on our servers, what you keep if you stop its fee is every template, draft and finding, as files.

No. Kent Applied doesn't resell any vendor's software and takes no referral fees, so no vendor pays us when you pick it. Our advice can point to a product you buy, one you already own, a tool we build or nothing at all. If something you already own will do the job, Tyler will tell you.

No. Policy drafts and committee charters are working documents for your counsel or compliance lead to review before you adopt them. Compliance calls stay with your compliance officer, and clinical decisions stay with your clinicians. No AI tool we build diagnoses, triages or recommends treatment. A predictive model scores risk, such as a likely no-show, for your staff to act on, and what to do about it stays their call.