Cross-silo systems strategist · Baraboo, Wisconsin

I read the whole system before I touch any part of it.

I work on public policy where it meets daily life: settlement dollars, housing, land, and the cost of new infrastructure. Right now I'm researching how Wisconsin can take on data centers and AI without passing the bill to households, farms and local budgets.

Why: the people who live with these decisions are usually the last to see the numbers.

1Dane County · opioid settlement structure
2Statewide · continuity map of 72 counties
3Statewide · data center and AI research and proposal
4Baraboo · housing data for the city
Patterns in the work

Different rooms, the same habits.

County policy, state emergency planning, city housing data, statewide energy policy. Here is what keeps showing up.

01

Read the rulebook closely

Statutes, eligibility rules and agency records usually hold the answer, or the gap. I go looking for both before I propose anything new.

Seen in: a gap in Wisconsin's emergency statute, and an existing Good Samaritan protection turned into a written agreement. Stop 1 →
02

Put the work where it can reach across

A good program in the wrong place can't see past its own department. Where a role sits decides what it can fix.

Seen in: arguing, over pushback, to place a new county role where it could work across departments. Stop 1 →
03

Count it, then say what the count can't tell you

I build the dataset when one doesn't exist, label every figure by where it came from, and write down the limits, including the limits of my own ideas.

Seen in: a statewide facility map, and a housing brief that names the trade-offs of the program I'm proposing. Stop 2 · Stop 5 →
04

Spend effort where it reaches the most people

Limited money and time go furthest where need is highest and the barrier is lowest to fix.

Seen in: simpler grants for small organizations, multilingual signs at the most-used gardens, and data center sites chosen where they lower local costs. Stop 4 →
What it adds up to: I'm most useful on problems that cross departments and belong to no single office, where the money, the rules and the people sit in different rooms.
The work, mapped

Five stops, 2025 to now.

Each one started with a problem nobody owned all of.

1
Dane County · Summer – Nov 2025

A $11.3 million fund needed a home that could reach every department.

Policy consultant (pro bono), Dane County Board, Opioid Settlement Subcommittee
Problem

The county was receiving opioid settlement money, $11.3 million so far, with no dedicated structure to plan and distribute it.

What I did
  • Wrote a white paper drawing on national practice, including a gap I found in Wisconsin's emergency-declaration statute (§ 323.11).
  • Recommended placing a new coordinator under the Department of Administration, despite pushback, so the role could work across departments.
  • Helped secure an agreement with the Sheriff and District Attorney for non-prosecution under existing Good Samaritan protections for NGOs doing street-level interventions.
  • Recommended a simplified grant process for small community organizations, and funding the role from interest on settlement funds.
What changed: the county adopted that structure in 2025 RES-274 and created a full-time Opiate Settlement Coordinator position during a hiring freeze.
2
Statewide · May – Aug 2025

Every court and DA facility in 72 counties, on one map.

Continuity planning intern, Wisconsin Department of Administration
Problem

Continuity planning needs to know where critical facilities are. The state's records didn't include coordinates.

What I did
  • Built a database of court and district attorney facilities across all 72 counties.
  • Added latitude and longitude to the records for the first time, so they could be mapped.
  • Standardized the records used in preparedness exercises, wrote templates for the state alert system, and took part in radiological emergency exercises.
3
Statewide · Sep 2026 – present

Who pays when data centers come to Wisconsin?

Independent policy research
Question

How can Wisconsin host data centers and adapt to AI without shifting the costs onto households, farmland and local budgets?

What I track
  • Utility rate decisions, federal transmission filings and state tax law, including the PSC's 2026 data center tariff and the state sales-tax exemption.
  • Where state law limits what towns and counties can ask of new development, from levy limits to conditions on local approvals.
  • The other side of the ledger: unemployment insurance, job retraining and farmland conservation.
4
Statewide · Nov 2026 · Draft for review

Moving Wisconsin Forward, Together

A policy proposal built from that research. Core ideas below; the full proposal is shared on request.
No bans, but full cost.

Development moves at the pace companies can pay their own way. Industry gets certainty. Communities get a real say.

01AI pays its shareWhen AI replaces work, the systems that work paid into stay funded, including schools and childcare.
02On Wisconsin's termsThe state leads on siting, projects carry their own grid costs, and neighbors can see the water and noise data.
03Land, water and farmsData center dollars fund voluntary restoration of flood-prone and streamside land. Productive farmland stays productive.
04Shared corridorsNew power lines follow the rail corridors that link Wisconsin's cities, moving freight off crowded highways.

What it does not do: no bans or moratoriums, no mandates on farmers, and no new agencies where an existing program already works.

5
Baraboo · Nov 2026 – present

My city's housing committee got the numbers, and the caveats.

Volunteer, Baraboo Community Development Authority housing committee

The committee is updating the city's housing needs assessment. I compiled public data from the Census, the city, the state and the tax roll into a five-page brief and offered it for the committee to verify and use.

46%
of renter households spend 30% or more of income on housing
~2%
rental vacancy, against the 5–7% the city's study calls healthy
~54%
of homes built before 1978, when lead paint was banned
Sources: City of Baraboo 2022 Housing Study; City permit counts via the draft 2040 Comprehensive Plan; U.S. Census Bureau ACS 2020–2024. Compiled Nov. 17, 2026.
How it's built: every figure is labeled by where it came from, and a section on limits notes that the acquisition-rehab model I'm developing adds no new units.
OfficialCounted from the tax rollEstimate, method statedNeeds city check
In development

What's next on the map.

Writing projects underway. Each builds on the work above.

Researching

Working-Class Policy Barrier Review

A long-form look at the Wisconsin laws that shape what working people can bargain for, from local preemption and project labor agreements to prevailing wage, Act 10 and right-to-work.

In progress

Critical Improvements

A working list of revisions to Moving Wisconsin Forward from a structured review. Each claim is checked against its source, weak points are fixed or cut, and every change is logged before it goes into the official version.

Scoping

The Weight We Carry

An analysis of favorability under current laws

A holistic look at people's rights under contracts and user agreements, their rights and responsibilities, and which laws favor corporations or place the burden on individuals. The goal is a weighted measure of how favorable current law is to individuals versus corporations.

Short essays from this work begin in January 2027.

How I think, in practice

Dutton's Intueri

A personal research collection that treats judgment like perception. Each "sense" is a separate model that measures one kind of signal. Intueri, Latin for "to gaze upon," is the act of combining them into a single read, without moralizing.

  • Weights are set before the results. No adjusting the model after seeing what it says.
  • Hard evidence outranks claims. Regulatory and adjudicated records come before self-reported or advocacy-framed sources, by a rule set in advance.
  • Physical measures come first. The senses closest to what's physically extracted carry the most weight. Reactive signals come after.
  • Toy models stay labeled as toy models until they survive testing.

Status: five of six senses drafted. Sight is in design. Combining them into one judgment is the open problem, and the most interesting one.

In design · untested

A look inside: agency and innovation

One piece of Intueri still being worked out, shown as it stands. It's being considered as part of the sight volume, Visus. It has not been tested.

Exhibit 1

Innovation grows likelier when everyone is included, regardless of credentials

Theory: as individual and collective agency converge, with all members of society counted, innovation becomes more likely. Conceptual model; hypothetical positions.

Research anchors

Horizontal axis (gating)Collins (1979)Credentials work as status closure and inflate over time, beyond certifying skill.
Left and right polesBandura (2006)Agency is exercised in personal, proxy and collective forms.
One continuous curveGiddens (1984)Structure is both medium and outcome of social practices.
Collective poleOstrom (1990)Groups can build self-governing institutions that avert the tragedy of the commons.
Vertical axisEmirbayer & Mische (1998)Agency is habitual or oriented toward alternative futures.
Each correspondence to this model is a proposal. None of these works tested the lemniscate.

Source: conceptual model by Eleanor Dutton. Literature anchors: Bandura (2006), Perspectives on Psychological Science 1(2); Giddens (1984), The Constitution of Society; Ostrom (1990), Governing the Commons; Emirbayer & Mische (1998), American Journal of Sociology 103(4); Collins (1979), The Credential Society.

Notes: Illustrative model, not an empirical result. The curve is the lemniscate of Bernoulli (a = 1) through the origin. Circle = individual agency point; square = collective agency point. Inclusion = 1 − mean(|x|)/a over the two points, so the origin is fully inclusive and a lobe tip is fully gated. Innovation index = exp(−(d/a)²) × inclusion, where d is the distance between the two points; a relative index, not a probability. Both rules were fixed before the examples were computed. The index does not tell progressive from regressive pairs (A and C score the same), so the regime is read from the axis. Positions are hypothetical.

Dutton's Intueri at work

How it could steer a legislative or fiscal decision

An illustrative use case. A state is weighing whether to keep, condition or change a tax exemption for a large new industry. Here is how Intueri would structure the read for lawmakers and planners.

1

Set the rules first

Weights, thresholds and a ranking of evidence are fixed before any data comes in, and published with the result.

Before measuring
2

Measure the physical

Land, water and habitat effects, from regulatory records first. These carry the most weight.

Smell · Touch
3

Read the signals

The lasting financial and social shadow: how reversible it is, who carries it, and for how long.

Hearing · Taste
4

Count what's closed off

Which future options each choice removes, such as farmland that can't come back.

Sixth sense
5

Check who's included

Whether the benefits reach people regardless of credentials, or stop at a gate.

Sight · in design
6

One combined read

A single judgment across all senses, with its uncertainty shown, not hidden.

Intueri

What decision-makers get

  • A side-by-side read of each option (keep, condition, phase in, or end) on the same predetermined scale.
  • The trade-offs that are hard to undo, named up front, so they aren't discovered after a vote.
  • Where a condition or a monitoring requirement would change the result, which points to what a bill or budget line should require.
  • A record of the rules used, so the analysis can be checked and repeated by others.

What it won't do

  • Hand down a moral verdict. It measures consequences; people decide.
  • Average away disagreement. When sources conflict, the rule set in advance decides which counts more.
  • Tune itself to a preferred answer. If the result is inconvenient, it is reported as it is.
  • Claim more certainty than it has. Untested parts are labeled as untested.

Illustrative use case only. No real project, company or bill has been scored. Parts of Intueri, including the sight volume and the step that combines the senses, are still in design.

Contact

Working on something that crosses departments?

I'd like to hear about it. Send a note, or find me on LinkedIn.

LinkedIn