Middle-market company
Something about the insurance program feels off.
Renewal strategy, claims pattern, broker process, contract requirement, captive idea, retention level, or total cost of risk.
See program reviewInsurance / risk / AI workflow / capital
Operating judgment, without the theater.
I write, build, invest, and advise across insurance, risk, AI workflow, and capital-backed growth. The common thread is simple: ask better questions before choosing the answer.
Send the renewal issue, broker concern, captive idea, claim pattern, contract requirement, AI workflow, or acquisition question. I'll tell you if there's something useful to do.
20+
years building, selling, running, and advising insurance businesses
50+
M&A transactions from founder, seller, operator, advisor, and diligence seats
Live
AI governance framework, technical workbench, and public GitHub proof
5
professional designations across insurance, risk, and coverage analysis
CPCU, CIC, ARM, CRM, TRIP
Start here
Choose the problem, not the label.
I work across three related situations. They overlap, but the starting point matters.
Investor or operator
The deal math needs operating truth.
Revenue quality, producer behavior, service capacity, retention risk, and where post-close drag can hide.
See capital lensInsurance or AI leader
AI is entering real insurance work before the controls are clear.
Use-case selection, review points, source traceability, audit trail, and governance before AI touches real insurance work.
See AI governanceThe frame
One person, four modes.
The connective tissue is pattern recognition: reading business situations quickly, saying what's actually true, and acting before the answer is obvious to everyone else.
Risk Advisory
Questions before market access.
For middle-market companies, I start by understanding the business, the current program, and the issue underneath the insurance request. Until that work is done, any recommendation would be premature.
- Signal
- Discovery before answers.
- Audience
- Owners, CFOs, controllers, risk managers.
Operating beliefs
Make the judgment visible.
Client work
Insurance work for middle-market companies.
Through Arvor Insurance, I work with middle-market companies as a commercial insurance and risk advisor.
The work starts with discovery. Until I understand the history, the facts, and where you're trying to take the business, I can't offer a useful solution. I want to understand how the business works, where risk actually shows up, what the current program is doing well, and what's sitting underneath the request.
Sometimes the answer is a better renewal strategy. Sometimes it's program redesign, higher retention, alternative risk transfer, captive feasibility, contractual review, or a different way to manage claims and risk throughout the year.
Process
Discovery comes before the answer.
First we understand the business. Then we decide whether the insurance program, risk strategy, or broker process needs to change.
Understand the business
Operations, contracts, claims, growth plans, capital structure, and the places risk actually shows up.
Pressure-test the current program
Coverage, retentions, claims handling, broker process, market strategy, and whether the structure still fits.
Build the plan
Program design, placement strategy, risk management priorities, and the first practical moves before renewal pressure takes over.
Where I help
- Current insurance program review
- Broker relationship review or oversight
- Risk management and total cost of risk
- Program design and placement strategy
- Deductible, loss-sensitive, and retention analysis
- Alternative risk transfer
- Captive feasibility and placement
- Contractual insurance review
- Renewal strategy and claims stewardship
Essays
Claims I can defend.
This should read like a working desk, not a post archive. The pieces below are active claims: useful enough to publish, concrete enough to defend, and open to revision as the work teaches more.
Brokerage is where insurance AI gets real.
Why it matters: producers and service teams touch the messy work that slide decks skip: intake, account prep, document review, follow-through, and renewal pressure.
Read brokerage essayAI governance has to name the human owner.
Why it matters: a policy is only useful if someone knows when to review, log, escalate, correct, or say no.
Read governance essayAn AI-native brokerage should make the senior broker more available.
Why it matters: the goal is more advisory time, cleaner account judgment, and fewer hours lost to mechanical assembly.
Read operating model essayThe real brokerage moat is renewal discipline.
Why it matters: retention is usually won months before the market sees the account, not in the last scramble.
Read renewal essayMost insurance problems are business problems first.
Why it matters: the policy is often where the cost shows up, but the cause usually sits in operations, contracts, claims, or growth.
Read business problems essayPrivate equity often buys production, then underestimates operating drag.
Why it matters: post-close results depend on service capacity, producer behavior, account quality, and what the P&L stops tolerating.
Read operating drag essayI don't even know if I can help you yet.
Why it matters: good insurance work starts by understanding the business before anyone offers the answer.
Read discovery essayThe best risk advisory starts before insurance is mentioned.
Why it matters: if the underlying risk gets better, the market conversation changes. Price follows the risk.
Read risk advisory essayOpen questions
The questions I keep coming back to.
These are the things I'm testing through writing, builds, insurance work, and conversations with people who have real operating stakes.
Capital lens
Where deal math meets operating truth.
The capital-backed growth lane isn't separate from the insurance work. It's the same pattern recognition applied to acquisition quality, producer economics, retention, and what happens after close.
Quality of revenue
How much of the book depends on a few producers, fragile client relationships, or renewal heroics nobody has written down?
Producer behavior
The model can look clean while the sales culture resists the exact discipline the model assumes.
Service capacity
Margin depends on what the service team can absorb before response time, documentation, and client trust start to slip.
Post-close truth
Integration risk usually shows up through retention, handoffs, claims, market relationships, and the accounts nobody wanted to inspect closely.
Workbench
Tools and experiments around the workflows I write about.
The workbench is public enough to show what I'm building, but not dressed up as a software company. The point is practical evidence.
Live / Governance
Insurance AI Governance Framework
Risk tiers, review expectations, controls, and practical boundaries for AI use in regulated insurance work.
- Problem
- Teams need usable AI rules before the risk becomes abstract.
- Does
- Maps use cases to review levels, controls, and owner decisions.
- Taught
- Governance gets useful when it fits the work people already do.
Technical / MCP
InsuranceXDate MCP
A TypeScript MCP wrapper built around messy carrier and filing data, with the workarounds documented instead of hidden.
- Problem
- Useful insurance data is fragmented, inconsistent, and hard to query cleanly.
- Does
- Exposes structured searches through a repeatable MCP interface.
- Taught
- The wrapper is only valuable if bad data states are visible.
In build / Documents
Insurance Document Assistant
Structured answers from policy excerpts, submissions, schedules, endorsements, and loss runs with sources that can be checked.
- Problem
- Teams lose time hunting for answers buried in insurance documents.
- Does
- Returns sourced answers from uploaded files for human review.
- Taught
- Trust depends on traceability more than speed.
Planned / Renewal prep
Broker Workflow Copilot
Account briefs, open items, and meeting prep generated from client inputs so the team edits instead of assembling from scratch.
- Problem
- Renewal prep still depends on too much manual assembly.
- Does
- Drafts the account brief, open items, and meeting prep path.
- Taught
- The best AI assist is the one a team can correct quickly.
Strategic advisory
Useful when the question needs judgment, not a packaged program.
I take on a small number of narrow advisory conversations where the question is real, the stakes are practical, and the answer is not already decided.
Good fit
- Insurance distribution strategy
- AI workflow and governance in regulated operations
- Producer operating models and retention economics
- Private equity diligence around brokerage quality
Bad fit
- Generic coaching
- Motivational founder content
- AI theater without operating ownership
- Projects where the answer is already decided
Contact
If the question is real, keep the note plain.
Send me 3-5 sentences on what's happening, what decision is coming up, what feels off, and what would make the conversation useful. If I think I can help, I'll tell you how. If not, I'll tell you that too.