AI Strategy for Emerging Managers and Small Funds

1. The Lean Fund's Real Constraint Is Hours, Not Headcount

An emerging manager runs the same job a large fund runs. Source deals, screen them, do diligence, write the memo, answer LPs, watch the portfolio. The difference is not the work. It is that two or three people do all of it.

So the binding constraint at a small fund is never headcount, because you cannot hire your way out of an early fund's economics. The constraint is hours. The partners are doing the sourcing and the spreading and the LP updates and the monitoring, often the same week, and the thing that gets dropped is whatever the calendar runs out of time for. A promising deal goes unscreened because nobody had the afternoon. That is the cost that does not show up on a budget.

This reframes the whole AI question for a lean team. The point is not to cut cost, because there is little to cut. The point is to buy back hours so the small team can cover more ground at the same headcount. That is a strategy worth having, and it is a different one from what a large firm needs. The general version is the AI strategy and roadmap; this is the lean-fund cut of it.

2. Why AI Tilts the Field Toward Small Teams

The usual assumption is that AI favors the big firm with the budget and the data team. For a lot of this work, the opposite is true.

A large fund's advantage has always been capacity: more analysts to read more CIMs, more associates to spread more borrowers, more people to staff more diligence. AI is capacity you can rent by the seat. It does not close the gap on relationships or capital or track record, and it should not pretend to. But on the throughput work, the reading and drafting and summarizing, a three-person team with good AI setups can now cover ground that used to require a much larger team. The capacity a big firm bought with payroll, a small firm can now rent with software.

There is a second edge that gets missed. A small team can actually change how it works. There is no committee, no legacy process, no twelve stakeholders to align. The partner decides on Monday and the workflow is different by Friday. Adoption, which is the thing that kills most large rollouts, is a near-non-issue at three people who all want it. Small is a disadvantage on capital and a real advantage on speed of change.

3. What a Sub-1-Billion-Dollar GP Should Do First

A lean fund should point AI at the three places where hours leak fastest, in roughly this order.

Sourcing and screening capacity. This is usually first, because it directly fixes the worst symptom: good deals going unlooked-at for lack of time. A fast first read of a CIM or teaser that pulls the key facts and flags the obvious concerns lets two partners screen the pipeline of a much larger team. IR and DDQs. An emerging manager spends a disproportionate share of its time raising and reporting, and a lot of that is drafting: LP updates, diligence questionnaires, data-room answers. AI drafting from your own prior materials, with a partner approving, gives back hours during exactly the stretch when there are none. Portfolio monitoring. Reading the reporting across companies and producing a short, consistent digest of what changed keeps a small team on top of the book without a dedicated person.

All three share the lean-fund logic: they buy back partner hours on assembly work, leaving the judgment where it belongs. Start with whichever one hurts most this quarter. For most early funds raising or deploying, that is sourcing or IR. Which one creates the most value for your specific firm is the subject of how much a fund should spend on AI and the value map below.

4. What to Ignore

As much strategy for a small fund is in what to ignore as in what to do. Most of the enterprise AI conversation is not for you, and chasing it wastes the one thing you are short of.

Ignore the firm-wide platform build. Ignore the six-figure custom system, the data lake, the dedicated AI engineering team, the model evaluations, the vendor bake-offs with a procurement committee. Those exist to solve coordination and scale problems that a three-person fund does not have. Ignore the pressure to have an opinion on every new model and feature, which is a full-time job that produces nothing for you. And ignore tools sold on a demo that does not map to a job you actually do every week.

The discipline is to treat almost the entire enterprise stack as not yet your problem. A small fund wins by doing a few high-value things with off-the-shelf tools, set up well, not by building what a large firm builds at a tenth of the scale. If you ever do need a build, it comes later, after a workflow has paid for itself, and the decision logic is in build, buy, or partner.

5. The Minimum Viable Stack

A lean fund does not need much. It needs four things, set up well. More than this, early, is usually waste.

This is achievable in weeks, not quarters, and most of it is configuration rather than engineering. The assistant and the knowledge base are paid subscriptions set up thoughtfully. The one built workflow is the highest-pain job turned into a repeatable setup. The governance is a page. A fund with these four is not behind the enterprise, it is ahead of most of it, because it has the parts that get used and none of the parts that gather dust.

6. Buy Before You Build

For a small fund the default is almost always buy, not build. Off-the-shelf tools, configured well, deliver most of the value at a fraction of the cost and none of the maintenance.

A custom build is a real commitment. It costs money you would rather deploy, and it costs something subtler: someone has to own it, update it, and fix it when a model changes underneath. A two-person fund does not have that someone. A good business-tier subscription, set up around your work, gets you most of the way and stays current without you. The cases where a build earns its keep, a workflow so specific and so heavily used that no tool fits, are rare at a small fund, and they arrive after the workflow has already proven itself on bought tools.

So the rule is simple: start by buying and configuring, prove the workflow, and only consider building the day a tool genuinely cannot do the job. Buying first is not the timid choice for a lean team, it is the correct one, and it keeps your scarce capital and attention on investing. The full framework for when that flips is build, buy, or partner.

7. Fractional Over Full-Time

A small fund cannot hire a head of AI, and it does not need one. What it needs is the expertise, not the salary, and that is what fractional gives you.

A full-time AI hire is the wrong shape for a lean team twice over. It is a fixed cost an early fund cannot justify, and the role is not full-time work at this size: setting up the stack and the workflows is a project, and keeping them current is a few days a month, not forty hours a week. Hiring full-time for a part-time job means overpaying for a person who is underused. The work is real and the full-time role is not.

Fractional leadership solves both. You get the person who has set this up before, who picks the right tools, builds the workflows, and keeps you current as the models change, for a fraction of a hire's cost and only as much time as the work needs. This is the head of AI you cannot afford yet, and it is exactly what an AI Operating Partner is. Whether to upskill the team, hire, or go fractional is laid out in who should own AI at your firm.

8. Governance a Small Team Can Actually Run

Governance sounds like a big-firm word, and small funds often skip it. That is a mistake, because LPs ask emerging managers the same AI questions they ask everyone, and an examiner does not grade on fund size.

The good news is that a small team needs a small version, and a small version is enough. One page, three answers. Which tools are approved for deal and LP material, and which are not. What you will tell an LP who asks how you use AI and protect their data, because they are asking now, in diligence. And what you will tell an examiner about how the firm supervises it. That is governance a two-person fund can write in an afternoon and actually follow, which is the whole point, because a policy nobody can run is not a policy.

The trap is the opposite of what small funds fear. The risk is not over-governing, it is having no answer at all when an LP asks, which reads as unserious during a raise. A one-page, honestly-followed rule beats a fifty-page document nobody reads, and it is enough to clear the bar at your size. If the firm faces a real exam or a sophisticated LP, the deeper version is AI governance.

9. The First 90 Days on a Small Budget

Here is the whole plan for a lean fund, on a budget that fits an early fund, run in three rough stages.

Weeks 1 to 2. Put the team on one good assistant, write the one-page data rule, and pick the single workflow that hurts most this quarter (usually screening or DDQs). Weeks 3 to 6. Build that one workflow well and run it on real work: the actual CIMs, the actual questionnaire. Stand up the shared knowledge base so the tool sounds like the firm. Measure the hours it gives back. Weeks 7 to 12. Once the first workflow is genuinely depended on, add the second from your priority list, and bring in fractional help if the setup is more than the team can carry alongside the day job.

The discipline is the same one that works at any size, just cheaper: one workflow won before the next, real work not demos, a number at the end. A lean fund can run this for the cost of a few subscriptions plus some focused outside help, and come out of a quarter covering meaningfully more ground at the same headcount. That is the return that matters for a small team, and it is reachable fast.

10. Where to Start

Pick the one job that cost you a deal or a weekend this quarter. Put the team on one good assistant, write the one-page data rule, and turn that job into a single repeatable workflow on real work. Ignore the rest of the enterprise stack for now.

Buy before you build, keep governance to a page, and get fractional help instead of a hire you cannot justify. A lean fund that does this well is not behind the big firms on the throughput work, it is ahead of most of them, because it can change how it works in a week and it spent nothing it did not need to.

If you want the lean-fund version built around your firm, an AI Readiness Sprint picks the workflow that buys back the most hours and stands it up on your real work, fast. We then act as the part-time head of AI you cannot hire yet, an AI Operating Partner who keeps your small stack sharp and current for a fraction of a full-time cost.