work with us
Scoped before work. Invoiced on request.
The core is MIT and stays free; these are the engagement modes around it. Fees are in USD, and every mode starts with a written scope – the number on day one is the number on the invoice unless the scope changes.
Discovery sprint
$2,500 fixed · one week
Map your tools, define the MCP surface, agree on auth and deployment. You leave with a written scope and a fixed quote for the build.
Governance readiness assessment
$15,000 fixed · 2–4 weeks
An AI-system inventory, a control crosswalk, and a prioritized remediation roadmap – the evidence a board or an auditor asks for first.
MCP integration build
from $15,000 · 2–4 weeks
One system connected to a hardened MCP server: auth, logging, a deployment recipe, and a runbook. Fixed scope, change orders only.
Governance & receipts integration
from $25,000
Dharma gates, receipt emission, and audit surfaces wired into your agent stack – verifiable claims instead of trust-me logs.
Private AI deployment
from $25,000
A local-first memory and governance deployment on your hardware: the wm stack, stores, and runbooks, with no required cloud path.
Research collaboration
from $25,000 · scoped
A specific question, a pre-registered method, and a public report at the end. Negative results are published as results.
Advisory retainer
$5,000/month
A pool of engineering hours, a response SLA, and quarterly roadmap reviews for teams running agents in production.
Consulting
$250/hour
Short, specific engagements: architecture review, evaluation design, or a second opinion before you commit to a build.
Founding collaborator: the first two engagements run at half list in exchange for a public case study and reference rights. Grants and foundations are welcome and do not buy outcomes. Start at /contact with a sentence about your system and the question you are trying to answer.