The work we have already done.
Different industries. The same method. Read these for the shape of the problem, because under the surface it is the one we solve everywhere.
Why there are no names here
Every engagement runs under NDA.
Our clients hand us the data they guard most closely, and keeping it quiet is part of what they pay for. We show you the work by what it did, never by whose it was. The discretion is the point, and it is the same discretion we would extend to you.
with
M.G.
Case studies
Different industries, the same method.
The organization needed to answer its funders with confidence. We reconciled its data into one verified source it now owns.
Public proof is certified aggregate figures only. No record-level data, and no personally identifiable information, appears here or anywhere on this site. The named, logo-bearing version of this case ships only on the organization's written consent; until then it is presented anonymized and without uncertified figures.
Twelve practice areas had grown twelve ways of documenting decisions. We defined ownership for the highest-risk ones, and the next review came back clean.
One analyst assembled the monthly report by hand, always right and always late. We separated the mechanical reconciliation from the human judgment. The read now runs in seconds.
Interactive tools
See the method work. Answer a few questions.
Executive
Your AI systems, the decisions they touch, and the people who own them. An interactive map with a drift timeline that shows how a personnel change today becomes a downstream failure six months out.
PE Firms
Ten companies, one operating-risk read. Status by company, the count of decisions nobody owns, and one urgent flag each. Drill into any name.
Banking
A board-committee AI risk report in ninety seconds. Risk inventory, accountability gaps, regulatory alignment across Fair Lending, ECOA, and BSA/AML, and a risk-ranked action plan.
Operations
Five questions, a scored read on whether your operation runs to a standard. It flags the two areas that need work first and writes the one-page report.
Nonprofit
How AI runs across your programs, fundraising, and operations. Decision ownership, beneficiary-data safeguards, and an operating-maturity read your board can act on.
Human-in-the-Loop
Watch a decision-authority map build itself as you answer. Five questions place each decision where it belongs and show who owns the call.
What these have in common
A line nobody had drawn.
A nonprofit, a financial firm, an operating business. Nothing alike on the surface. Every one came down to the same thing. Skilled people doing by hand what a system should carry, and a line no one had drawn for what the system was allowed to run on its own. We drew the line. Then we built.
Why Mesa exists
Why Mesa exists.
We built Mesa on one conviction. Automation works when authority is clear. Name who owns each decision, set the line a machine will not cross, and AI becomes the most dependable part of your operation. That is the work we do, and it is the work we do first.
The people
A principal is directly engaged on every engagement. Everything else runs as bounded, checked automation.
Dan Dollins
Managing Principal
Dan Dollins is the Architect of Mesa Group, the firm that maps how decisions get made inside a business and where AI can safely take them, then builds only what belongs. He designed Mesa to be the one firm that does both, mapping and building, answerable to no software vendor, so every recommendation serves the client's business rather than a vendor's roadmap.
Jay Shurling
Principal, Business Development
The build work is led by a senior developer with deep range, from a single operational tool to a multi-system build. Forward-deployed when the work calls for it.
The standard doesn't flex
The same rigor, whatever the size of your business.
A ten-person company gets the rigor we would bring to the largest client we can imagine. The tone of the conversation meets you where you are. The quality underneath does not change.