Capital
The thesis assumed margin. The portfolio company still runs on email.
A newly hired COO is asked to improve margins without adding people at the same rate as revenue. The work that would create that leverage is still a diligence folder, a monthly pack, and a process nobody has mapped.
How I see it
PE value creation is an operations problem wearing a deal jacket
The company is under pressure. A COO has a clock. The operating model cannot be replaced, but one expensive workflow can.
Two layers matter. At the firm: diligence, IC memos, portfolio reporting, and follow-up across companies. At the portfolio company: the same coordination tax every other upper-middle-market operator has, now with a hold-period clock.
AI is useful when it shortens the packet and the operating loop. It is not useful as a firm-wide transformation story. The first win should be a workflow a partner or a COO can take into the next operating review.
Miami and South Florida have a dense set of PE-backed distribution, services, real estate, hospitality, and specialty-finance companies. The work is still one process with a baseline.
Common mistakes
What teams usually get wrong.
Funding a firm chatbot instead of a portfolio workflow
A knowledge assistant for the deal team can help. It does not create the margin the thesis promised.
Asking every portfolio company for an AI roadmap
A survey of ideas is not value creation. Pick one company and one expensive process.
Confusing reporting speed with operating leverage
A faster monthly pack is useful. Changing the workflow that produces the numbers is the larger prize.
A useful diagnostic
Five questions before you fund the work.
Is the pressure on the firm packet or on a portfolio-company process?
Those are different owners, different systems, and different first bets.Which workflow would change the operating review if it got 25 percent faster or cheaper?
If the list is a sector thesis, you do not have a workflow yet.Does the COO or operating partner have authority to retire steps?
Without that authority, AI becomes another workstream in the 100-day plan.Can you baseline the current path in 30 days?
If not, the hold period will be spent on discovery theater.What result would make the next board meeting easier to walk into?
If the answer is a vision deck, you are funding narrative, not leverage.
Economic model
Hold-period testannual workflow value × years remaining in the hold × confidence the owner can change the process
If the third term is near zero, wait. A strong spreadsheet will not ship without an operator.
Three credible paths
How far should you go?
Do not force one solution. Choose the path the economics, the risk, and the organization can support.
Improve the work you already have
Keep the process mostly intact and use AI on the bottlenecks that create delay, rework, or follow-up.
The workflow is already sound and a few steps create most of the friction.
Gains are usually incremental. The operating economics do not change much.
Redesign the workflow around AI
Question every handoff, queue, and duplicate step, then rebuild the process around what AI can now do.
The process grew over years and coordination now costs more than the work itself.
Requires process change, clearer ownership, and a willingness to retire old steps.
Put an agent on a high-value outcome
Give an agent responsibility for one valuable result across systems, with humans at the control points that require judgment, authority, or risk acceptance.
The workflow is high-value, variable, multi-step, and worth engineering for production.
Needs stronger architecture, evaluation, controls, and monitoring. This is not a prompt project.
When this is the wrong next step
Do not start with a sector program.
- The request is a firm-wide AI vision for the annual meeting.
- No portfolio company will give access to the people who run the work.
- The only metric is that the firm appears modern.
- The process is a one-off deal situation with no repeatable volume.
A useful next step
Bring the workflow the hold period is waiting on.
We will separate firm packets from portfolio operations and name a first bet a COO can own.
Discuss an AI opportunity