Kent Applied is an applied AI and data science consultancy. We design AI strategy, build data platforms and AI-enabled applications, and put them into production. Our deepest expertise is healthcare, where the data is regulated and mistakes are expensive.
Most organizations don't need another AI vendor. They need the strategy, the data foundation, and the working systems, built to fit how they operate. Kent Applied delivers all three.
Where AI genuinely pays in your organization, what it costs, what to skip, and the roadmap to get there. Vendor evaluation included; hype filtered out.
Predictive modeling, statistical analysis, self-service analytics, and decision support your leadership uses.
Modern stacks stood up from zero: warehousing, pipelines, semantic layers, governance. The foundation everything else depends on.
Document processing, internal copilots, autonomous agents, and workflow automation, taken from concept to production.
Healthcare data is regulated, messy, and political, and it punishes generalists. Kent Applied was built here: HIPAA-governed Medicare claims research that supported federal drug-pricing policy analyses, clinical trial programming for FDA submissions, behavioral health analytics, and FERPA-compliant architecture.
We also treat healthcare delivery as what it is: a flow problem. Constrained capacity, queues, and variability respond to operations science, and that lens, rare in health tech, is behind our signature Throughput Diagnostic.
Engagements are structured around HIPAA and PHI boundaries from day one, with a BAA in place where needed. Early-stage work is designed to avoid PHI entirely.
years of healthcare data and analytics behind the practice
AI products designed and shipped to production
annual client renewal at a healthcare SaaS company running the client-health scoring we built
regulatory regimes worked under: HIPAA, CMS, FDA, FERPA
Fixed prices and written scope, never an hourly meter. Six are set offers; the seventh is scoped around your problem. See how an engagement works →
A structured assessment of your data and workflows: where AI genuinely pays in your organization, what it costs, and what to skip. You get a costed roadmap, a ranked opportunity analysis, and one working proof of concept.
We find your binding constraint, measure it from your own data, put a dollar figure on every day it persists, and deliver a relief plan with a projected throughput gain. Operations science applied to patient flow, discharge delays, scheduling, and admin bottlenecks. For hospitals in CMS's mandatory TEAM model, the diagnostic produces the 30-day episode cost and flow baseline for each of the five surgery categories.
A scoped automation or data system, built, tested, documented, and handed off with training: intake and document processing, reporting pipelines, internal copilots, data architecture. Written acceptance criteria define "done" before work starts.
Ongoing AI and data leadership one to two days a week: roadmap ownership, vendor evaluation, build oversight, and training your team until they can run it without us.
Live sessions tuned to each audience: executive briefings, hands-on workshops for daily users, and first-steps classes for teams that are new or skeptical. Real prompts on your real workflows, guardrails your compliance office can sign, and a playbook your team keeps. Taught by someone with 20+ AI adoption sessions inside a national healthcare organization and years of guest lectures in a graduate clinical informatics course at the University of San Diego.
Your shortlist of AI vendors, scored against your actual requirements: capability on your real cases, security and HIPAA posture, what happens to your data, integration cost, and the build-versus-buy math. You get a scored comparison, a recommendation with the reasoning shown, and negotiation points for the contract.
Some problems don't fit a menu. Bring yours: a dataset nobody can query, a model that needs an owner, a process that quietly eats a hire's worth of hours. If AI and data science can solve it, we scope it, write the acceptance criteria, and price it fixed. If they can't, you'll hear that on the first call, for free.
Three AI systems our principal designed and built to power a public data platform on healthcare and aging in America: one finds the insights, one turns them into visuals, one guards the quality. They're built as one interoperable system, not three separate tools. Together they show what production AI looks like in practice.
An autonomous agent that scans the public web daily for emerging healthcare topics and turns them into publication-ready insight candidates.
A self-serve engine that converts written insights into branded, publication-ready infographics, without a design bottleneck.
An automated QA system that continuously audits a live public data platform and reports to every audience that needs to know.
Five commitments, written into every statement of work.
Nothing is finished until it's been shown working on your real cases.
Acceptance criteria go in the statement of work before anything starts. You know the price and the definition of done up front.
Engagements are structured around your data boundaries from day one. Every statement of work states plainly how and whether AI tools touch your data.
Documentation and training are deliverables. The goal is your team running the system on their own.
If the problem doesn't need AI, we'll say so on the first call and point you at the cheaper fix.
Short, sourced pieces on healthcare bottlenecks, AI that earns its keep, and the operations science behind both.
Boarding is measured in the ED and handed to the ED to fix. The constraint is inpatient capacity and discharge timing.
Hospitals eat $2,000 to $4,000 for every day a discharge-ready patient stays in a bed. The fix is a queueing problem, not a software feature.
Health systems report $3.20 back per dollar invested in AI, but the returns cluster in a few unglamorous places. Here's the map.
Data science and AI executive. Thirteen years in healthcare data, published in JAMA Network Open, with Medicare claims research that supported federal drug-pricing policy analyses and production AI products shipped end to end.
Bring the problem you're trying to solve. In twenty minutes you'll get an honest read on whether it's an AI problem, a data problem, or neither.
Or email [email protected], or connect on LinkedIn.