A collection of problems I've had the privilege of messing with.
Exploring what becomes possible when software suddenly gets cheaper to create.
My AI work spans product development, rapid prototyping, AI-assisted software engineering, data workflows, curriculum tools, analytics, and experimental applications.
The common thread isn't “put AI in everything.”
It's identifying places where AI fundamentally changes what is practical to build.
I've used modern AI systems as development partners—moving between product concept, interface design, architecture, code, testing, and iteration at speeds that would have seemed slightly irresponsible a few years ago.
Now it's Tuesday.
What if AI acted more like a mirror than an oracle?
AIReflective is an independent exploration of conversational AI for reflection and journaling.
Rather than positioning AI as an expert dispensing answers, the concept explores how an AI system can ask useful questions, help someone capture their thinking, and surface patterns across time.
It combines my interests in AI, product design, privacy, data, and human-computer interaction.
Turning organizational data into organizational awareness.
As Director of Business Intelligence at the Tennessee Department of Health, my work expanded well beyond traditional reporting.
I worked across strategic planning, organizational performance, executive decision support, program and funding visibility, and enterprise initiatives designed to help a large public-sector organization better understand itself.
Projects included strategic-plan performance systems, executive dashboards, program inventories, cross-departmental reporting, funding and program alignment, and tools designed to turn fragmented information into something leaders could actually act on.
The technical work mattered. But the harder problem was almost always organizational.
Strategy eventually has to survive contact with Monday morning.
My public-sector work grew from business intelligence into broader organizational strategy and operations.
I've worked on enterprise strategic planning, executive leadership structures, performance management, organizational design, operational efficiency, major funding initiatives, team development, and cross-functional programs involving dozens of stakeholders.
This is work I particularly enjoy because it sits between ideas and execution.
Strategy asks where we're going. Operations asks whether anyone packed the car. You need both.
Building for environments where reality gets a vote.
My work in education technology spans AI strategy, software development, product experimentation, analytics, curriculum tooling, and internal systems.
Projects have included district usage analytics, AI-assisted curriculum planning, data applications, internal AI adoption, rapid product prototyping, and exploring how AI changes both what we build and how we build it.
Education is a useful antidote to technological arrogance. A product eventually has to work for a real teacher, in a real classroom, with real students, under real constraints.
The demo doesn't get a vote. Reality does.
Data is only useful if someone can figure out what it's trying to tell them.
This project focused on transforming education-product usage data into an experience district administrators could actually use.
The challenge wasn't collecting more information. It was turning enormous amounts of existing information into answers.
Who is using the product? Where? How often? What patterns matter? What should someone pay attention to?
That's product design disguised as analytics.
Making the blank page less expensive.
Curriculum development involves constant iteration: lessons move, structures change, requirements collide, and ideas have to become concrete before anyone can decide whether they're good.
I built AI-assisted tooling to help curriculum teams rapidly generate, organize, manipulate, and evaluate scope-and-sequence structures.
AI wasn't replacing the curriculum expert. It was giving the expert something to push against.
Different organizations. Familiar problems.
My consulting work has given me the opportunity to move between organizations, industries, and problem types without assuming that the solution from the last room belongs in the next one.
I've advised and worked with organizations around technology, data, operations, strategy, systems, and implementation.
Consulting reinforced a principle that has become central to how I work:
Diagnose before you prescribe.
People understandably arrive asking for solutions. A dashboard. A system. Automation. AI. A new process.
But solutions are expensive ways to discover you've misunderstood the problem.
The first job is figuring out what's actually happening.
Jason Matthew Howes