Build with Me

Built by someone who has run these operations.

I have been the finance person, the ops director, and the development support worker inside movement organizations — not in passing, but across eighteen years and the full range of structures movement organizations build — fiscal sponsorships, dual-entity operations, multi-funder grant portfolios, and the rest. From that ground I build MIAgents (Movement Intelligence Agents) scoped to the functions I have actually lived inside: financial operations, fundraising, institutional memory, program reporting. This is not general AI consulting. It is a practitioner building tools for a sector the practitioner knows from the inside, for organizations whose data stays theirs.

Where this work fits

The conditions insider advantage matters most.

This is not a general AI consulting offer. It is specifically for organizations navigating movement-sector financial and operational complexity. The practitioner background pays off in four conditions in particular.

Fiscal sponsorship. A sponsor relationship means reconciliation between the sponsor's reports and your internal books lives in two systems that rarely speak to each other, and finance staff carries the gap by hand. Tools designed from the inside start there instead of pretending the gap does not exist.
Dual-entity C3/C4 compliance. Staff time allocation across entities, lobbying expenditures against applicable limits, and the line between advocacy and political campaign intervention. Every transaction touches compliance. An outside vendor has to be taught this. A practitioner already knows where the decisions get made and where they break.
Multi-year grant complexity. Accrual schedules that do not match cash timing, payment pipelines stacking up against reporting deadlines, compliance obligations that carry forward across fiscal years. This is where the finance seat holds the most untranslated strategic intelligence in the organization.
Knowledge loss from staff turnover. When a finance manager leaves or a development coordinator moves on, the organization loses the relationships, the pattern recognition, and the quiet decisions that were never documented. This is where institutional memory tooling pays for itself.

How we work together

Three ways to build.

Tier 1

MIAgent Scoping

A 30-minute initial conversation to gauge what your organization is carrying and determine whether your own MIAgent is the right fit. We talk through the work, the constraints, and where intelligence is currently getting lost — and decide together whether there is something worth scoping further.

Book a scoping session

Tier 2

MIAgent Build

I build, deploy, and document a scoped MIAgent for your organization. Financial ops and development MIAgents are the current catalog. Each build includes a training session for the staff who will use it, and documentation written in plain language. Your organization owns the artifact — code, data, and documentation. I retain generalizable architectural patterns (how to structure a grant-tracking agent, how to design a reconciliation workflow) but never retain your organization's data, configurations, or domain-specific knowledge.

Start with a scoping session

Tier 3

MIAgent Stewardship

I host, maintain, and iterate on your MIAgent on a monthly basis. Versioned updates as your organization's context evolves: new programs, new funders, new staff. For organizations that want the intelligence without the ongoing overhead of maintaining it themselves.

Let's talk about fit

Your data does not train external models.

Every agent I build runs through private API infrastructure with your organizational data separated from external systems. Every agent runs through API providers whose terms prohibit training on your data, with architectural separation that keeps your knowledge base isolated. Network-isolated deployment is available for organizations that require it.

Private API infrastructure

Every agent call routes through a private API. Your data layer stays under your control. Your data never touches a public endpoint or a shared index. Deploy on serverless edge, private cloud, or on-prem — no vendor lock-in. Network-isolated deployment with locally hosted models available on request.

Your data, your control

Organization-specific knowledge bases are isolated per client. No shared indexes, no cross-contamination between contexts. What your team knows stays inside your walls — never pooled, never reused for another engagement.

Authenticity as a design principle

Does the output sound like something the community would say, or something a consultant would say about the community? That question shapes how every agent is prompted, tested, and refined — a design standard, not an automated filter.

Bring a problem, not a product spec.

The best first step is a scoping conversation. Bring the movement-sector complexity your organization is carrying: the fiscal sponsor reconciliation that never quite balances, the compliance burden that has outgrown spreadsheet tracking, the strategic intelligence held by one or two people on staff that would walk out the door with them. We will figure out together whether a MIAgent is the right tool, what it would need to know, and what done looks like for your organization.

Book an exploration session

No commitment. 30 minutes. You leave with clarity.